api-20260821T200735Z / page 3 of 5 / documentation capture

build-with-claude/vision-coordinates Changed · +31 / -1 lines

## Turn resizing into an error with `transformations`

from line 444

444444 The [Token counting](https://platform.claude.com/docs/en/build-with-claude/token-counting) endpoint estimates an image's token cost from its dimensions without fully processing it, so a successful count doesn't mean the image is within the Messages API's [request limits](https://platform.claude.com/docs/en/build-with-claude/vision#request-limits). An image can count successfully and still be rejected when you send it.

445445</Note>

446446

447## Turn resizing into an error with `transformations`

448

449Pre-resizing only protects your coordinates while your pipeline keeps producing the right sizes. A new image source or a switch to a model on a different resolution tier can quietly reintroduce server-side resizing. To turn that silent drift into a visible error, set the optional `transformations` field on an image content block in a [Messages](https://platform.claude.com/docs/en/api/messages) request:

450

451```json

452{

453 "type": "image",

454 "source": { "type": "base64", "media_type": "image/png", "data": "..." },

455 "transformations": { "oversized_image": "error" }

456}

457```

458

459A request whose marked image (any block that sets `"oversized_image": "error"`) would be resized is rejected with a 400 `invalid_request_error` naming the image's dimensions and the largest dimensions that fit. Whether an image triggers the rejection depends on the [limits of every model the request names](https://platform.claude.com/docs/en/build-with-claude/vision-coordinates#how-claude-resizes-and-pads-images): the 1920×1080 example below is rejected by a standard-tier model but fits within the high-resolution tier:

460

461```text wrap

462messages.0.content.0: image dimensions 1920x1080 exceed the maximum image size of a model named on this request and would be downsized to 1456x819; scale the image to at most 1456x819 or set the image's oversized_image setting to "downsize"

463```

464

465Rescale to the reported target and resend: the target is the largest size, at your image's aspect ratio, that every model named on the request accepts. How marked images interact with the [server-side fallback beta](https://platform.claude.com/docs/en/build-with-claude/refusals-and-fallback#server-side-fallback) is described with that feature; in every mode, a marked image is never served resized.

466

467The setting is per image. `"oversized_image": "downsize"` (the default when the field is omitted) keeps automatic resizing as described on this page. Each image block is checked only against its own setting, so one request can mix images whose dimensions are load-bearing (a screenshot you'll click on) with images where resizing is harmless (a logo). What the setting does and does not change:

468

469* Padding ([which never discards content](https://platform.claude.com/docs/en/build-with-claude/vision-coordinates#how-claude-resizes-and-pads-images)), format conversion, and orientation correction proceed as usual.

470* The hard limits (8000 px on the longest side, and the stricter per-image limit on [many-image requests](https://platform.claude.com/docs/en/build-with-claude/vision#request-limits)) are separate rejections; this setting never lets an image past them.

471* Images supplied by URL or file ID are checked once their bytes have been fetched; those rejections carry the same message without the leading position, so they don't identify which image failed; only embedded base64 images are named by position in the error.

472* [PDF pages](https://platform.claude.com/docs/en/build-with-claude/pdf-support) are rasterized server-side at dimensions you don't control; the `document` block does not accept the field (an image block nested inside a document's content accepts it like any other).

473* A marked image whose dimensions cannot be determined is rejected rather than passed through: that rejection reports that the image's source dimensions could not be determined, not the resize message quoted above. No image that sets `"error"` reaches the model resized.

474

475The [Token counting](https://platform.claude.com/docs/en/build-with-claude/token-counting) endpoint honors `transformations` too, rejecting an embedded image exactly as the Messages API would, so you can check whether an embedded image fits without being resized, before running inference. Counting never fetches images supplied by URL or file ID, so a marked image from those sources is checked only at Messages time, as described above.

476

447477## Rescale coordinates when you cannot pre-resize

448478

449If you cannot pre-resize (for example, when the image comes from an upstream system you can't modify), use the resize helper from [Resize your image before uploading](https://platform.claude.com/docs/en/build-with-claude/vision-coordinates#resize-your-image-before-uploading) to recover the dimensions Claude saw, then map the coordinates Claude returns into normalized coordinates or back onto your original image. Claude resizes oversized images rather than rejecting them, up to the API's [request limits](https://platform.claude.com/docs/en/build-with-claude/vision#request-limits). Beyond those limits the request fails with a validation error instead. Pass the tier limits that match the model you called: the wrong tier's limits recover the wrong resized dimensions and silently shift every coordinate. This approach requires knowing the pixel dimensions of the image you uploaded, so it does not apply to PDF uploads.

479If you cannot pre-resize (for example, when the image comes from an upstream system you can't modify), use the resize helper from [Resize your image before uploading](https://platform.claude.com/docs/en/build-with-claude/vision-coordinates#resize-your-image-before-uploading) to recover the dimensions Claude saw, then map the coordinates Claude returns into normalized coordinates or back onto your original image. Unless an image [opts into an error instead](https://platform.claude.com/docs/en/build-with-claude/vision-coordinates#oversized-image-error), Claude resizes oversized images rather than rejecting them, up to the API's [request limits](https://platform.claude.com/docs/en/build-with-claude/vision#request-limits). Beyond those limits the request fails with a validation error instead. Pass the tier limits that match the model you called: the wrong tier's limits recover the wrong resized dimensions and silently shift every coordinate. This approach requires knowing the pixel dimensions of the image you uploaded, so it does not apply to PDF uploads.

450480

451481Screenshots and zoom images that you return to the [computer use](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool#handle-coordinate-scaling-for-higher-resolutions) and [browser use](https://platform.claude.com/docs/en/agents-and-tools/tool-use/browser-use-tool#targets-and-coordinates) toolsets are an exception to automatic resizing. The API rejects a `tool_result` image that exceeds the model's limits with a validation error instead of resizing it. Resize those images in your application before returning them, then scale the coordinates Claude returns back to your screen's dimensions.

452482

claude_api_primer Changed · +3 / -4 lines

from line 181

181181 ```python Python

182182 import anthropic

183183 import base64

184 import httpx

184 import httpx2

185185

186186 # Option 1: Base64-encoded image

187187 image_url = "https://platform.claude.com/docs/images/vision-example.jpg"

188188 image_media_type = "image/jpeg"

189 image_data = base64.standard_b64encode(httpx.get(image_url).content).decode("utf-8")

189 image_data = base64.standard_b64encode(httpx2.get(image_url).content).decode("utf-8")

190190

191191 message = anthropic.Anthropic().messages.create(

192192 model="claude-opus-5",

from line 262

262262

263263<CodeGroup exclude="shell:cURL, typescript, csharp, go, java, php, ruby">

264264 ```bash CLI

265 ant messages create \

266 --transform content --format yaml <<'YAML'

265 ant messages create --transform content --format yaml <<'YAML'

267266 model: claude-opus-5

268267 max_tokens: 16000

269268 thinking:

cli-sdks-libraries/sdks/python Changed · +20 / -18 lines

from line 36

3636

3737## Requirements

3838

39Python 3.9 or later is required.

39Python 3.10 or later is required. If you are upgrading from a 0.x release of the SDK, see the [v1 migration guide](https://github.com/anthropics/anthropic-sdk-python/blob/main/MIGRATION.md) for the list of breaking changes.

4040

4141## Usage

4242

from line 102

102102

103103### Using aiohttp for better concurrency

104104

105For improved async performance, you can use the `aiohttp` HTTP backend instead of the default `httpx`:

105For improved async performance, you can use the `aiohttp` HTTP backend instead of the default `httpx2`:

106106

107107```python

108108import os

from line 362

362362 )

363363except anthropic.APIConnectionError as e:

364364 print("The server could not be reached")

365 print(e.__cause__) # an underlying Exception, likely raised within httpx

365 print(e.__cause__) # an underlying Exception, likely raised within httpx2

366366except anthropic.RateLimitError as e:

367367 print("A 429 status code was received; we should back off a bit.")

368368except anthropic.APIStatusError as e:

from line 426

426426

427427## Timeouts

428428

429By default requests time out after 10 minutes. You can configure this with a `timeout` option, which accepts a float or an `httpx.Timeout` object:

429By default requests time out after 10 minutes. You can configure this with a `timeout` option, which accepts a float or an `httpx2.Timeout` object:

430430

431431```python

432import httpx

432import httpx2

433433from anthropic import Anthropic

434434

435435# Configure the default for all requests:

from line 439

439439

440440# More granular control:

441441client = Anthropic(

442 timeout=httpx.Timeout(60.0, read=5.0, write=10.0, connect=2.0),

442 timeout=httpx2.Timeout(60.0, read=5.0, write=10.0, connect=2.0),

443443)

444444

445445# Override per-request:

from line 588

588588

589589### Accessing raw response data (for example, headers)

590590

591The "raw" `Response` returned by `httpx` can be accessed through the `.with_raw_response` property on the client. This is useful for accessing response headers or other metadata:

591The "raw" `Response` returned by `httpx2` can be accessed through the `.with_raw_response` property on the client. This is useful for accessing response headers or other metadata:

592592

593593```python

594594client = Anthropic()

from line 606

606606print(message.content)

607607```

608608

609These methods return an `APIResponse` object.

609These methods return an `APIResponse` object. On the async client they return an `AsyncAPIResponse`, and `.parse()`, `.read()`, `.text()`, and `.json()` must be awaited.

610610

611611### Streaming response body

612612

from line 645

645645To make requests to undocumented endpoints, you can use `client.get`, `client.post`, and other HTTP verbs. Options on the client, such as retries, are respected when making these requests.

646646

647647```python

648import httpx

648import httpx2

649649

650650response = client.post(

651651 "/foo",

652 cast_to=httpx.Response,

652 cast_to=httpx2.Response,

653653 body={"my_param": True},

654654)

655655

from line 670

670670

671671### Configuring the HTTP client

672672

673You can directly override the [httpx client](https://www.python-httpx.org/api/#client) to customize it for your use case, including support for proxies and transports:

673The SDK sends requests with [httpx2](https://httpx2.pydantic.dev), an API-compatible fork of `httpx`. To customize the HTTP client, including proxies and transports, pass your own [httpx2 client](https://httpx2.pydantic.dev/api/#client) as `http_client`:

674674

675675```python

676import httpx

676import httpx2

677677from anthropic import Anthropic, DefaultHttpxClient

678678

679679client = Anthropic(

from line 681

681681 base_url="http://my.test.server.example.com:8083",

682682 http_client=DefaultHttpxClient(

683683 proxy="http://my.test.proxy.example.com",

684 transport=httpx.HTTPTransport(local_address="0.0.0.0"),

684 transport=httpx2.HTTPTransport(local_address="0.0.0.0"),

685685 ),

686686)

687687```

from line 693

693693```

694694

695695<Note>

696 Use `DefaultHttpxClient` and `DefaultAsyncHttpxClient` instead of raw `httpx.Client` and `httpx.AsyncClient` to ensure the SDK's default configuration (such as timeouts and connection limits) is preserved.

696 Use `DefaultHttpxClient` and `DefaultAsyncHttpxClient` instead of raw `httpx2.Client` and `httpx2.AsyncClient` to ensure the SDK's default configuration (such as timeouts and connection limits) is preserved. The `http_client` argument must be an `httpx2` client. Passing a client from the separate `httpx` package raises a `TypeError`.

697697</Note>

698

699Tracing and mocking tools that patch `httpx` itself, such as OpenTelemetry's `HTTPXClientInstrumentor`, Sentry's `httpx` integration, `respx`, or `pytest-httpx`, do not see the SDK's requests by default. To use them, call `httpx2.alias_httpx()` once at startup, before anything imports `httpx`. This makes `import httpx` resolve to `httpx2` for the whole process.

698700

699701### Managing HTTP resources

700702

manage-claude/compliance-activity-feed Changed · +4 / -2 lines

from line 14

1414 Both Compliance Access Keys (`sk-ant-api01-...`) carrying this scope and Admin API keys (`sk-ant-admin01-...`) can call the Activity Feed. See [Set up the Compliance API](https://platform.claude.com/docs/en/manage-claude/compliance-api-access) for the conditions under which each key type carries the scope.

1515</Check>

1616

17The Activity Feed records every authentication, chat, file, project, administrative, and platform action that occurs in your organization, in reverse chronological order. Activities are queryable within 1 minute of occurring and are retained for 6 years.

17The Activity Feed records authentication, chat, file, project, administrative, and platform activity across your organization and returns it in reverse chronological order. Activities are queryable within 1 minute of occurring and are retained for 6 years.

1818

1919```bash cURL

2020curl --fail-with-body -sS \

from line 150

150150| ---------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------- |

151151| `user_actor` | A signed-in claude.ai or Claude Console user took the action. | `email_address`, `user_id`, `ip_address`, `user_agent` |

152152| `api_actor` | A request called the Claude API or the Compliance API with a customer-issued API key. Compliance API calls produce this actor type for both Compliance Access Keys and Admin API keys. | `api_key_id`, `ip_address`, `user_agent` |

153| `admin_api_key_actor` | An organization admin used an Admin API key to manage users, invites, workspaces, or API keys. | `admin_api_key_id` |

153| `admin_api_key_actor` | An organization admin used an Admin API key to manage users, invites, workspaces, or API keys. | `admin_api_key_id`, `ip_address`, `user_agent` |

154154| `unauthenticated_user_actor` | An action occurred before sign-in completed, for example `sso_login_initiated`. | `unauthenticated_email_address`, `ip_address`, `user_agent` |

155155| `anthropic_actor` | Anthropic acted on the organization, for example through internal tooling. | `email_address` (always `null`; present for shape consistency with `user_actor`, because Anthropic operators are not represented by individual email) |

156156| `scim_directory_sync_actor` | An identity provider (such as Okta, Microsoft Entra ID, or JumpCloud) pushed a change through SCIM directory sync. | `workos_event_id`, `directory_id`, `idp_connection_type` (nullable; for example `OktaSCIMV2`, `AzureSCIMV2`) |

157

158A `claude_*_viewed` activity means a Claude app loaded content, not that a person viewed it. Types such as `claude_chat_viewed`, `claude_file_viewed`, and `claude_project_viewed` are recorded each time a Claude app loads the chat, file, or project from Anthropic's servers. Repeated loads are not deduplicated. The web, desktop, and mobile apps load content at different moments, sometimes in the background, and can display a cached copy without loading it. Counts of these activities vary by platform as a result, and they do not correspond to messages sent or screens viewed.

157159

158160<Note>

159161 **Build forward-compatible handlers.** Pass through unrecognized `type` and `actor.type` values, and ignore fields your handler does not expect, so your integration keeps working when new activity types ship.

manage-claude/compliance-integration-patterns Changed · +10 / -7 lines

from line 94

9494

9595Each `Activity` carries fields you can join against events already in your SIEM (Splunk, Datadog, Microsoft Sentinel, Cribl, or similar):

9696

97| Compliance API field | Join target |

98| --------------------- | ----------------------------------------------- |

99| `actor.user_id` | Your identity provider's stable user identifier |

100| `actor.email_address` | Directory email when a stable ID is unavailable |

101| `actor.ip_address` | Network, VPN, and endpoint logs |

102| `created_at` | Time-window correlation across any source |

97| Compliance API field | Join target |

98| --------------------- | ----------------------------------------------------------------------- |

99| `actor.user_id` | Your identity provider's stable user identifier |

100| `actor.email_address` | Directory email when a stable ID is unavailable |

101| `actor.ip_address` | Network, VPN, and endpoint logs |

102| `actor.user_agent` | Endpoint and device inventory, and the client app that made the request |

103| `created_at` | Time-window correlation across any source |

103104

104`actor.user_id` and `actor.email_address` are present when `actor.type` is `user_actor`; check the discriminator before reading them. `user_id` is a stable, opaque identifier for the user account: it is consistent across every Compliance API endpoint and activity payload, and it does not change when the user's email or display name changes. Use `user_id`, not `email_address`, as the primary join key.

105`actor.user_id` and `actor.email_address` are present when `actor.type` is `user_actor`. `actor.ip_address` and `actor.user_agent` are absent on some actor types, such as `anthropic_actor` and `scim_directory_sync_actor`. Check the discriminator before reading any of these fields. `user_id` is a stable, opaque identifier for the user account: it is consistent across every Compliance API endpoint and activity payload, and it does not change when the user's email or display name changes. Use `user_id`, not `email_address`, as the primary join key.

105106

106107Calls to the Compliance API itself emit `compliance_api_accessed` activities. Ingest these alongside other activity types so your SIEM records who queried compliance data, and when. Pass `activity_types[]=compliance_api_accessed` to scope the query, then in your client, read `actor.api_key_id` from each activity whose `actor.type` is `api_actor` to attribute the access to a specific Compliance Access Key or Admin API key.

107108

from line 137

136137* The starting cursor and the terminal `last_id`.

137138* The number of records exported.

138139* The run timestamp and the `request-id` of the final page.

140

141Activity volume is not a completeness check. The `claude_*_viewed` activity types, such as `claude_chat_viewed`, follow each app's loading pattern (see [Understand the Activity object](https://platform.claude.com/docs/en/manage-claude/compliance-activity-feed#understand-the-activity-object)). A period with chat messages but no `claude_chat_viewed` activities does not on its own indicate missing data. Rely on the traversal and the overlap or reconciliation pass described in [Window polling](https://platform.claude.com/docs/en/manage-claude/compliance-integration-patterns#window-polling) instead.

139142

140143The content endpoints (chats, files, projects, project attachments, and local and remote session transcripts) serve Claude Enterprise data only. The Activity Feed surfaces administrative and resource events organization-wide. The Compliance API does not include:

141144

managed-agents/cloud-sandboxes-reference Changed · +37 / -25 lines

### Document and media processing ### Browser automation

from line 14

1414

1515## Programming languages

1616

17| Language | Version | Package manager |

18| -------- | ------- | --------------- |

19| Python | 3.12+ | pip, uv |

20| Node.js | 20+ | npm, yarn, pnpm |

21| Go | 1.22+ | go modules |

22| Rust | 1.77+ | cargo |

23| Java | 21+ | maven, gradle |

24| Ruby | 3.3+ | bundler, gem |

25| PHP | 8.3+ | composer |

26| C/C++ | GCC 13+ | make, cmake |

17| Language | Version | Package manager |

18| -------- | --------------------------- | -------------------- |

19| Python | 3.10, 3.11, 3.12, and 3.13 | pip, uv, poetry |

20| Node.js | 20, 21, and 22 (default) | npm, yarn, pnpm, bun |

21| Go | 1.24 (default) and 1.25 | go modules |

22| Rust | Stable toolchain (rustup) | cargo |

23| Java | OpenJDK 21 | maven, gradle |

24| Ruby | 3.1, 3.2, and 3.3 (default) | bundler, gem |

25| PHP | 8.3 | composer |

26| C/C++ | GCC 13 and Clang | make, cmake, ninja |

2727

28Common Python data and document libraries, including NumPy, pandas, Matplotlib, openpyxl, python-docx, python-pptx, and pypdf, are installed for the `python3` interpreter.

29

2830## Databases

2931

30| Database | Description |

31| ----------------- | -------------------------------------------------- |

32| SQLite | Pre-installed, available immediately |

33| PostgreSQL client | `psql` client for connecting to external databases |

34| Redis client | `redis-cli` for connecting to external instances |

32| Database | Description |

33| ------------- | ----------------------------------------------------------------------------- |

34| PostgreSQL 16 | Server and `psql` client are installed. The server is not running by default. |

35| Redis 7 | Server and `redis-cli` are installed. The server is not running by default. |

36| SQLite | Available through language bindings, such as Python's `sqlite3` module. |

3537

36<Note>

37 Database servers (such as PostgreSQL and Redis) are not running in the sandbox by default. The sandbox includes client tools for connecting to external database instances. SQLite is fully available for local use.

38</Note>

39

4038## Utilities

4139

4240### System tools

from line 41

4341

4442* `git` - Version control

4543* `curl`, `wget` - HTTP clients

46* `jq` - JSON processing

44* `jq`, `yq` - JSON and YAML processing

4745* `tar`, `zip`, `unzip` - Archive tools

48* `ssh`, `scp` - Remote access (requires a networking mode that allows the destination host)

49* `tmux`, `screen` - Terminal multiplexers

46* `tmux` - Terminal multiplexer

5047

5148### Development tools

5249

from line 50

5350* `make`, `cmake` - Build systems

5451* `docker` - Container management (limited availability)

5552* `ripgrep` (`rg`) - Fast file search

56* `tree` - Directory visualization

57* `htop` - Process monitoring

5853

5954### Text processing

6055

from line 57

6257* `vim`, `nano` - Text editors

6358* `diff`, `patch` - File comparison

6459

60### Document and media processing

61

62* `ffmpeg` - Audio and video processing

63* ImageMagick (`convert`, `identify`) - Image manipulation

64* `pandoc` - Document conversion

65* LibreOffice (headless) - Office document conversion

66* Poppler utilities (`pdftotext`, `pdftoppm`) and `qpdf` - PDF processing

67* `tesseract` - Optical character recognition (English language data)

68* TeX Live (`pdflatex`, `xelatex`, `latexmk`) - Typesetting

69

70### Browser automation

71

72* Playwright (Python and Node.js) - Browser automation library

73* Chromium (`/opt/pw-browsers/chromium`) - Browser used by Playwright, not on `PATH`

74

75The sandbox sets `PLAYWRIGHT_BROWSERS_PATH` to `/opt/pw-browsers`, so the pre-installed Playwright packages find Chromium there without configuration. The Python package is installed for the `python3` interpreter. Use the pre-installed packages rather than installing another Playwright version, which would look for a browser build that is not present. Firefox and WebKit are not installed.

76

6577## Sandbox specifications

6678

6779| Property | Value |

6880| ---------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |

69| Operating system | Ubuntu 22.04 LTS |

81| Operating system | Ubuntu 24.04 LTS |

7082| Architecture | x86\_64 (amd64) |

7183| Memory | Up to 8 GB |

7284| Disk space | Up to 10 GB |

managed-agents/define-outcomes Changed · +7 / -11 lines

from line 238

238238 ```

239239

240240 ```php PHP

241 // The PHP SDK exposes the Files API under the beta namespace; field names can differ from other SDKs.

242241 use Anthropic\Client;

243242 use Anthropic\Core\FileParam;

244243

from line 255

256255 MD;

257256 file_put_contents('/tmp/rubric.md', $rubricText);

258257

259 $rubric = $client->beta->files->upload(

258 $rubric = $client->files->upload(

260259 file: FileParam::fromResource(fopen('/tmp/rubric.md', 'r'), contentType: 'text/markdown'),

261260 );

262261 echo "Uploaded rubric: {$rubric->id}\n";

from line 335

336335 --transform id --raw-output)

337336

338337 # Define the outcome — agent starts working on receipt

339 ant beta:sessions:events send \

340 --session-id "$SESSION_ID" <<YAML

338 ant beta:sessions:events send --session-id "$SESSION_ID" <<YAML

341339 events:

342340 - type: user.define_outcome

343341 description: Build a DCF model for Costco in .xlsx

from line 710

712710

713711## Retrieve deliverables

714712

715The agent writes output files to `/mnt/session/outputs/` inside the sandbox. Once the session is idle, fetch them through the [Files API](https://platform.claude.com/docs/en/build-with-claude/files) scoped to the session. Filtering by `scope_id` requires the `managed-agents-2026-04-01` beta header on the list request, so the SDK and CLI examples make that call through the `beta` namespace and pass the header explicitly.

713The agent writes output files to `/mnt/session/outputs/` inside the sandbox. To retrieve them, list files through the [Files API](https://platform.claude.com/docs/en/build-with-claude/files) with the session ID as the `scope_id`, then download them by ID. Filtering by `scope_id` requires the `managed-agents-2026-04-01` beta header on the list request, so the SDK and CLI examples make that call through the `beta` namespace and pass the header explicitly. Files appear in the list shortly after the agent finishes writing them, sometimes a few seconds after the session goes idle. If a file you expect is not listed yet, list again after a short delay; once it appears in the list, its upload has finished.

716714

717715<CodeGroup>

718716 ```bash cURL

from line 736

738736 ```bash CLI

739737 # List files produced by this session

740738 # scope_id filtering requires the managed-agents beta on the files request

741 ant beta:files list --scope-id "$SESSION_ID" \

742 --beta managed-agents-2026-04-01

739 ant beta:files list --scope-id "$SESSION_ID" --beta managed-agents-2026-04-01

743740

744741 # Download a file

745742 FILE_ID=$(ant beta:files list --scope-id "$SESSION_ID" \

from line 855

858855 ```

859856

860857 ```php PHP

861 // The PHP SDK exposes the Files API under the beta namespace; field names can differ from other SDKs.

862858 // List files produced by this session

863859 // scope_id filtering requires the managed-agents beta on the files request

864860 $files = $client->beta->files->list(scopeID: $session->id, betas: ['managed-agents-2026-04-01']);

865 foreach ($files->data as $file) {

861 foreach ($files->getItems() as $file) {

866862 echo "{$file->id} {$file->filename}\n";

867863 }

868864

869865 // Download a file

870 if (count($files->data) > 0) {

871 $content = $client->beta->files->download($files->data[0]->id);

866 if (count($files->getItems()) > 0) {

867 $content = $client->files->download($files->getItems()[0]->id);

872868 file_put_contents('/tmp/output.txt', $content);

873869 }

874870 ```

managed-agents/environments Changed · +5 / -7 lines

from line 161

161161 ```

162162

163163 ```bash CLI

164 ant beta:sessions create \

165 --agent "$AGENT_ID" \

166 --environment-id "$ENVIRONMENT_ID"

164 ant beta:sessions create --agent "$AGENT_ID" --environment-id "$ENVIRONMENT_ID"

167165 ```

168166

169167 ```python Python

from line 388

390388

391389| Field | Package manager | Example |

392390| ------- | ------------------------- | ------------------------------------------- |

393| `apt` | System packages (apt-get) | `"ffmpeg"` |

394| `cargo` | Rust (cargo) | `"[email protected]"` |

391| `apt` | System packages (apt-get) | `"graphviz"` |

392| `cargo` | Rust (cargo) | `"[email protected]"` |

395393| `gem` | Ruby (gem) | `"rails:7.1.0"` |

396394| `go` | Go modules | `"golang.org/x/tools/cmd/goimports@latest"` |

397395| `npm` | Node.js (npm) | `"[email protected]"` |

398| `pip` | Python (pip) | `"pandas==2.2.0"` |

396| `pip` | Python (pip) | `"sqlalchemy==2.0.30"` |

399397

400398### Networking

401399

from line 720

722720

723721## Pre-installed runtimes

724722

725Cloud sandboxes include common runtimes out of the box. See [Cloud sandbox reference](https://platform.claude.com/docs/en/managed-agents/cloud-sandboxes-reference) for the full list of pre-installed languages, databases, and utilities.

723Cloud sandboxes include common language runtimes, databases, and command-line tools out of the box. See [Cloud sandbox reference](https://platform.claude.com/docs/en/managed-agents/cloud-sandboxes-reference) for the full list.

726724

727725## Next steps

728726

managed-agents/files Changed · +15 / -14 lines

from line 24

2424 ```

2525

2626 ```bash CLI

27 FILE_ID=$(ant files upload \

28 --file data.csv \

29 --transform id --raw-output)

27 FILE_ID=$(ant files upload --file data.csv --transform id --raw-output)

3028 ```

3129

3230 ```python Python

from line 69

7169 ```

7270

7371 ```php PHP

74 // The PHP SDK exposes the Files API under the beta namespace; field names can differ from other SDKs.

75 $file = $client->beta->files->upload(

76 FileParam::fromResource(fopen($csvPath, 'r'), filename: 'data.csv', contentType: 'text/csv'),

72 $file = $client->files->upload(

73 file: FileParam::fromResource(fopen($csvPath, 'r'), filename: 'data.csv', contentType: 'text/csv'),

7774 );

7875 echo "File ID: {$file->id}\n";

7976 ```

from line 314

317314

318315 ```php PHP

319316 $resources = [

320 ['type' => 'file', 'file_id' => 'file_abc123', 'mount_path' => '/data.csv'],

321 ['type' => 'file', 'file_id' => 'file_def456', 'mount_path' => '/config.json'],

322 ['type' => 'file', 'file_id' => 'file_ghi789', 'mount_path' => '/src/main.py'],

317 ['type' => 'file', 'fileID' => 'file_abc123', 'mountPath' => '/data.csv'],

318 ['type' => 'file', 'fileID' => 'file_def456', 'mountPath' => '/config.json'],

319 ['type' => 'file', 'fileID' => 'file_ghi789', 'mountPath' => '/src/main.py'],

323320 ];

324321 ```

325322

from line 535

538535

539536## Listing and downloading session files

540537

541Use the [Files API](https://platform.claude.com/docs/en/build-with-claude/files) to list files scoped to a session and download them. Filtering by `scope_id` requires the `managed-agents-2026-04-01` beta header, so the list examples use the `beta` files namespace and pass that header explicitly.

538Use the [Files API](https://platform.claude.com/docs/en/build-with-claude/files) to list files scoped to a session and download them. Files the agent writes to `/mnt/session/outputs/` appear in the list shortly after the agent finishes writing them, sometimes a few seconds after the session goes idle. If an output file you expect is missing, list again after a short delay; once it appears in the list, its upload has finished.

542539

540Filtering by `scope_id` requires the `managed-agents-2026-04-01` beta header, so the list examples use the `beta` files namespace and pass that header explicitly.

541

543542<CodeGroup>

544543 ```bash cURL

545544 # List files associated with a session

from line 556

557556

558557 ```bash CLI

559558 # List files associated with a session

560 ant beta:files list --scope-id sesn_abc123 \

561 --beta managed-agents-2026-04-01

559 ant beta:files list --scope-id sesn_abc123 --beta managed-agents-2026-04-01

562560

563561 # Download a file

564562 ant files download --file-id "$FILE_ID" --output output.txt

from line 649

651649 ```

652650

653651 ```php PHP

654 // The PHP SDK exposes the Files API under the beta namespace; field names can differ from other SDKs.

655652 // List files associated with a session

656653 $files = $client->beta->files->list(

657654 scopeID: 'sesn_abc123',

658655 betas: ['managed-agents-2026-04-01'],

659656 );

657 foreach ($files->getItems() as $file) {

658 echo "{$file->id} {$file->filename}\n";

659 }

660660

661661 // Download a file

662 $content = $client->beta->files->download($files->data[0]->id);

662 $content = $client->files->download($files->getItems()[0]->id);

663663 file_put_contents('output.txt', $content);

664664 ```

665665

from line 696

696696* If you omit `mount_path`, the file is placed at `/mnt/session/uploads/<file_id>`

697697* Parent directories are created automatically

698698* Paths should be absolute (starting with `/`)

699* Files the agent writes to `/mnt/session/outputs/` become available through the Files API, scoped to the session; see [Listing and downloading session files](https://platform.claude.com/docs/en/managed-agents/files#listing-and-downloading-session-files)

699700

managed-agents/github Changed · +10 / -4 lines

from line 389

389389 ```

390390</CodeGroup>

391391

392The `resources[].authorization_token` authenticates the repository clone operation and is not echoed in API responses.

392A `github_repository` resource accepts the following fields:

393393

394| Field | Description |

395| --------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |

396| `type` | Required. Must be `"github_repository"`. |

397| `url` | Required. The repository's HTTPS URL in the form `https://github.com/<owner>/<repo>`, without a `.git` suffix. Other forms, including SSH URLs, are rejected with an `invalid_request_error`. |

398| `authorization_token` | Required. The GitHub token used to clone the repository. It is not echoed in API responses. See [Token permissions](https://platform.claude.com/docs/en/managed-agents/github#token-permissions). |

399| `mount_path` | Optional. The directory under `/workspace` to clone the repository into. Defaults to `/workspace/<repo-name>`. |

400| `checkout` | Optional. A branch (`{"type": "branch", "name": "main"}`) or commit (`{"type": "commit", "sha": "..."}`) to check out. Defaults to the repository's default branch. |

401

394402Mounting a repository also loads any skills stored in its root `.claude/skills` directory. Skills are discovered once per session, from the repository state checked out at session start. See [Load skills from a GitHub repository](https://platform.claude.com/docs/en/managed-agents/skills#load-skills-from-a-github-repository).

395403

396404## Token permissions

from line 753

745753 ```

746754

747755 ```bash CLI

748 ant beta:sessions:events send \

749 --session-id "$SESSION_ID" \

750 > /dev/null <<'EOF'

756 ant beta:sessions:events send --session-id "$SESSION_ID" > /dev/null <<'EOF'

751757 events:

752758 - type: user.message

753759 content:

managed-agents/session-operations Changed · +5 / -5 lines

from line 29

2929

3030The semantics of a `tools` or `mcp_servers` update are full replacement: the provided array is the new value. To preserve existing entries, `GET` the session, modify the array, and `POST` it back.

3131

32The session must be `idle` to update the agent. [Interrupt](https://platform.claude.com/docs/en/managed-agents/events-and-streaming#integrating-events) the session if you need to update the agent while it's running.

32The session must be `idle` to update the agent. To update the agent while the session is running, send a [`user.interrupt` event](https://platform.claude.com/docs/en/managed-agents/events-and-streaming#integrating-events) by itself and wait for the session to become `idle`.

3333

3434<CodeGroup defaultLanguage="CLI">

3535 ```bash cURL

from line 532

532532

533533## Archiving a session

534534

535Archive a session to prevent new events from being sent while preserving its history. A `running` session cannot be archived; send an [interrupt event](https://platform.claude.com/docs/en/managed-agents/events-and-streaming#integrating-events) if you need to archive it immediately.

535Archive a session to prevent new events from being sent while preserving its history. A `running` session cannot be archived; to archive one, send a [`user.interrupt` event](https://platform.claude.com/docs/en/managed-agents/events-and-streaming#integrating-events) by itself and wait for the session to become `idle`.

536536

537537<CodeGroup defaultLanguage="CLI">

538538 ```bash cURL

from line 581

581581

582582## Deleting a session

583583

584Delete a session to permanently remove its record, events, and associated sandbox. A `running` session cannot be deleted; send an [interrupt event](https://platform.claude.com/docs/en/managed-agents/events-and-streaming#integrating-events) if you need to delete it immediately.

584Delete a session to permanently remove its record, events, and associated sandbox. A `running` session cannot be deleted; to delete one, send a [`user.interrupt` event](https://platform.claude.com/docs/en/managed-agents/events-and-streaming#integrating-events) by itself and wait for the session to become `idle`.

585585

586Memory stores, vaults, skills, environments, and agents are independent resources and are not affected by session deletion. Files you uploaded through the Files API are also unaffected, but files the session itself produced are scoped to it and are permanently deleted along with its filesystem. Download anything you need to keep before deleting the session.

586Memory stores, vaults, skills, environments, and agents are independent resources and are not affected by session deletion. Files you uploaded through the Files API are also unaffected, but files the session itself produced are scoped to it and are permanently deleted along with its filesystem. Download anything you need to keep before deleting the session. An output file written at the end of the last turn can take a few seconds after the session goes idle to appear in the [session's file list](https://platform.claude.com/docs/en/managed-agents/files#listing-and-downloading-session-files), so check that the files you expect are listed first.

587587

588588<CodeGroup defaultLanguage="CLI">

589589 ```bash cURL

release-notes/overview Changed · +29 / -25 lines

### August 20, 2026

from line 12

1212 For updates to Claude Code, see the [complete CHANGELOG.md](https://github.com/anthropics/claude-code/blob/main/CHANGELOG.md) in the `claude-code` repository.

1313</Tip>

1414

15### August 20, 2026

16

17* We've released **v1.0 of the [Python SDK](https://platform.claude.com/docs/en/cli-sdks-libraries/sdks/python)**. The SDK's HTTP layer moves from `httpx` to [httpx2](https://httpx2.pydantic.dev), a maintained, API-compatible fork: build custom `http_client`, `Timeout`, and transport objects from `httpx2` (the `DefaultHttpxClient` helpers are unchanged), and call `httpx2.alias_httpx()` at startup if you rely on tracing or mocking libraries that patch `httpx`. v1.0 requires Python 3.10 or later and removes long-deprecated surface, including the legacy Text Completions API, the `temperature`, `top_p`, and `top_k` parameters on Messages methods, and the tool runner's client-side `compaction_control`. On the async client, `.with_raw_response` results now need `await response.parse()`, and `AnthropicBedrock` now raises an error when no AWS region is configured instead of defaulting to `us-east-1`. See the [v1 migration guide](https://github.com/anthropics/anthropic-sdk-python/blob/main/MIGRATION.md) for every change with before-and-after snippets.

18

1519### August 19, 2026

1620

17* The [computer use tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool) is now generally available on the Claude API as the `computer_toolset_20260801` toolset: no beta header, batch actions (several actions in one turn), `zoom` enabled by default, and per-member configuration through `configs`. Earlier beta versions remain available. Upgrading an existing integration changes the request shape and tool handling; see [Migrate from `computer_20251124`](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool#migrate-from-computer-20251124).

21* The [computer use tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool) is out of beta on the Claude API as the `computer_toolset_20260801` toolset: no beta header, batch actions (several actions in one turn), `zoom` enabled by default, and per-member configuration through `configs`. Earlier beta versions remain available. Upgrading an existing integration changes the request shape and tool handling; see [Migrate from `computer_20251124`](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool#migrate-from-computer-20251124).

1822* We've launched the [browser use tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/browser-use-tool) (`browser_toolset_20260801`), a client toolset for driving a browser that your application hosts. It works inside a browser viewport rather than a whole desktop, reading the page itself (its accessibility tree, elements, forms, and tabs) and adding element references, form input, tab management, download reporting, and opt-in file upload on top of screenshot-and-click control.

1923* Both toolsets are available for Claude Fable 5, Claude Mythos 5, Claude Opus 5, Claude Sonnet 5, and Claude Opus 4.8 on the Claude API.

20* The [Files API](https://platform.claude.com/docs/en/build-with-claude/files) is now generally available on the Claude API. Requests to the `/v1/files` endpoints, and Messages API requests that reference an uploaded file, no longer require the `files-api-2025-04-14` beta header. Requests sent without the header use the GA response format: [file expiration](https://platform.claude.com/docs/en/build-with-claude/files#file-expiration) (set `expires_in_seconds` when you upload a file; file objects report `expires_at`), and `page` and `next_page` [pagination](https://platform.claude.com/docs/en/api/overview#pagination) plus an `ids[]` filter when you [list files](https://platform.claude.com/docs/en/build-with-claude/files#list-files). `/v1/files` requests that still send the beta header keep working and return the previous response format.

21* [Agent Skills](https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview) and the Skills API (`/v1/skills`) are now generally available on the Claude API. Requests no longer require the `skills-2025-10-02` beta header, including Messages API requests that load Skills through the `container` parameter. Requests that still send the header continue to work unchanged. See [Using Agent Skills with the API](https://platform.claude.com/docs/en/build-with-claude/skills-guide).

22* The [Admin API](https://platform.claude.com/docs/en/api/admin) user-management endpoints for **Claude Enterprise** (claude.ai) organizations (members, invites, groups, and custom roles) are now generally available. The `anthropic-beta: ce-user-management-2026-07-13` header is no longer required on group and custom-role requests; requests that still send it are accepted unchanged. See [User management](https://platform.claude.com/docs/en/manage-claude/user-management).

24* The [Files API](https://platform.claude.com/docs/en/build-with-claude/files) is out of beta on the Claude API. Requests to the `/v1/files` endpoints, and Messages API requests that reference an uploaded file, no longer require the `files-api-2025-04-14` beta header. Requests sent without the header use the current response format: [file expiration](https://platform.claude.com/docs/en/build-with-claude/files#file-expiration) (set `expires_in_seconds` when you upload a file; file objects report `expires_at`), and `page` and `next_page` [pagination](https://platform.claude.com/docs/en/api/overview#pagination) plus an `ids[]` filter when you [list files](https://platform.claude.com/docs/en/build-with-claude/files#list-files). `/v1/files` requests that still send the beta header keep working and return the previous response format.

25* [Agent Skills](https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview) and the Skills API (`/v1/skills`) are out of beta on the Claude API. Requests no longer require the `skills-2025-10-02` beta header, including Messages API requests that load Skills through the `container` parameter. Requests that still send the header continue to work unchanged. See [Using Agent Skills with the API](https://platform.claude.com/docs/en/build-with-claude/skills-guide).

26* The [Admin API](https://platform.claude.com/docs/en/api/admin) user-management endpoints for **Claude Enterprise** (claude.ai) organizations (members, invites, groups, and custom roles) are out of beta. The `anthropic-beta: ce-user-management-2026-07-13` header is no longer required on group and custom-role requests; requests that still send it are accepted unchanged. See [User management](https://platform.claude.com/docs/en/manage-claude/user-management).

2327* You can now restrict which sites a Claude Managed Agents agent's `web_search` and `web_fetch` tools can reach. Set `allowed_domains` or `blocked_domains` on the tool's entry in the `agent_toolset_20260401` `configs` array; `web_fetch` also accepts `max_content_tokens` and `web_search` accepts `user_location`. Each `configs` entry is identified by its `name` and typed by an optional `type`, and requests that pass only `name`, `enabled`, and `permission_policy` continue to work; in the typed SDKs, `configs` entries become per-tool types. See [Restrict web search and web fetch domains](https://platform.claude.com/docs/en/managed-agents/tools#restrict-web-search-and-web-fetch-domains).

2428* Claude Managed Agents sessions that run in a [self-hosted sandbox](https://platform.claude.com/docs/en/managed-agents/self-hosted-sandboxes) can now attach [memory stores](https://platform.claude.com/docs/en/managed-agents/memory). The Python, TypeScript, and Go SDK workers download each attached store into the sandbox at its `mount_path` and sync the agent's changes back to the store. See [Use memory stores](https://platform.claude.com/docs/en/managed-agents/self-hosted-sandboxes#use-memory-stores).

2529* The session viewer in the Claude Console has been redesigned with a timeline minimap, a transcript grouped by model request, and an Inspector panel for session details and cost, raw events, per-tool statistics, mounted resources, and per-thread activity. See [Console observability](https://platform.claude.com/docs/en/managed-agents/events-and-streaming#console-observability).

from line 63

5963

6064### July 24, 2026

6165

62* We've launched **Claude Opus 5** (`claude-opus-5`), a step-change improvement over Claude Opus 4.8. Claude Opus 5 supports a [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) (both the default and the maximum), 128k max output tokens, and [thinking](https://platform.claude.com/docs/en/build-with-claude/thinking) on by default, at $5 / $25 USD per MTok, the same pricing as Claude Opus 4.8. It's available on the Claude API, [Claude in Amazon Bedrock](https://platform.claude.com/docs/en/build-with-claude/claude-in-amazon-bedrock), [Claude on Google Cloud](https://platform.claude.com/docs/en/build-with-claude/claude-on-vertex-ai), and [Claude in Microsoft Foundry](https://platform.claude.com/docs/en/build-with-claude/claude-in-microsoft-foundry). See [What's new in Claude Opus 5](https://platform.claude.com/docs/en/about-claude/models/whats-new-opus-5) for new features, behavior changes, and migration guidance, and the [models overview](https://platform.claude.com/docs/en/about-claude/models/overview) for complete specs.

66* We've launched **Claude Opus 5** (`claude-opus-5`), a step-change improvement over Claude Opus 4.8. Claude Opus 5 supports a [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) (both the default and the maximum), 128k max output tokens, and [thinking](https://platform.claude.com/docs/en/build-with-claude/thinking) on by default, at $5 / $25 USD per MTok, the same pricing as Claude Opus 4.8. It's available on the Claude API, [Claude in Amazon Bedrock](https://platform.claude.com/docs/en/build-with-claude/claude-in-amazon-bedrock), [Claude Platform on AWS](https://platform.claude.com/docs/en/build-with-claude/claude-platform-on-aws), [Claude on Google Cloud](https://platform.claude.com/docs/en/build-with-claude/claude-on-vertex-ai), and [Claude in Microsoft Foundry](https://platform.claude.com/docs/en/build-with-claude/claude-in-microsoft-foundry). See [What's new in Claude Opus 5](https://platform.claude.com/docs/en/about-claude/models/whats-new-opus-5) for new features, behavior changes, and migration guidance, and the [models overview](https://platform.claude.com/docs/en/about-claude/models/overview) for complete specs.

6367* On Claude Opus 5, disabling thinking is allowed only at effort `high` or below: `thinking: {"type": "disabled"}` with effort `xhigh` or `max` returns a 400 error, a breaking change from Claude Opus 4.8. See [What's new in Claude Opus 5](https://platform.claude.com/docs/en/about-claude/models/whats-new-opus-5#behavior-changes).

6468* [Effort](https://platform.claude.com/docs/en/build-with-claude/effort) is the primary control for steering Claude Opus 5: the model supports the full ladder (`low`, `medium`, `high`, `xhigh`, `max`), with `max` for capability-critical work.

6569* Mid-conversation tool changes are now in beta on Claude Fable 5, Claude Mythos 5, Claude Opus 4.8, and Claude Opus 5: add or remove tools between turns of a conversation while preserving the prompt cache. Include the `mid-conversation-tool-changes-2026-07-01` beta header in your requests.

from line 181

177181

178182### May 28, 2026

179183

180* We've launched **Claude Opus 4.8** (claude-opus-4-8), our most capable generally available model. Claude Opus 4.8 supports a [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) by default on the Claude API, Amazon Bedrock, Google Cloud, and Microsoft Foundry, 128k max output tokens, and the same set of tools and platform features as Claude Opus 4.7. See the [migration guide](https://platform.claude.com/docs/en/about-claude/models/migration-guide) for baseline settings, features, and migration guidance.

184* We've launched **Claude Opus 4.8** (claude-opus-4-8), our most capable widely released model. Claude Opus 4.8 supports a [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) by default on the Claude API, Amazon Bedrock, Google Cloud, and Microsoft Foundry, 128k max output tokens, and the same set of tools and platform features as Claude Opus 4.7. See the [migration guide](https://platform.claude.com/docs/en/about-claude/models/migration-guide) for baseline settings, features, and migration guidance.

181185* We've launched [mid-conversation system messages](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages). On Claude Opus 4.8, you can send `role: "system"` messages after a user turn (subject to [placement rules](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages#limitations)) in the `messages` array, preserving prompt cache hits when instructions change during a long-running session. No beta header is required.

182186* The [`stop_details`](https://platform.claude.com/docs/en/build-with-claude/refusals-and-fallback#refusal-response) field on refusal responses is now publicly documented; it returns a `category` (`cyber`, `bio`, or `null`) and a human-readable `explanation`, so your application can route different classes of refusal to the right next step. No beta header is required.

183187* On Claude Opus 4.8, the [effort parameter](https://platform.claude.com/docs/en/build-with-claude/effort) defaults to `high` across all surfaces, including Claude Code and the Messages API.

from line 236

232236

233237### May 4, 2026

234238

235* [Workload Identity Federation](https://platform.claude.com/docs/en/manage-claude/workload-identity-federation) is now generally available. Authenticate workloads to the Claude API with short-lived OIDC tokens from your own identity provider (AWS IAM, Google Cloud, GitHub Actions, Kubernetes, Microsoft Entra ID, Okta, SPIFFE, and more) instead of long-lived static API keys. Configure issuers and federation rules in the Claude Console, and the SDK handles token exchange and refresh automatically. See [Authentication](https://platform.claude.com/docs/en/manage-claude/authentication).

239* We've launched [Workload Identity Federation](https://platform.claude.com/docs/en/manage-claude/workload-identity-federation). Authenticate workloads to the Claude API with short-lived OIDC tokens from your own identity provider (AWS IAM, Google Cloud, GitHub Actions, Kubernetes, Microsoft Entra ID, Okta, SPIFFE, and more) instead of long-lived static API keys. Configure issuers and federation rules in the Claude Console, and the SDK handles token exchange and refresh automatically. See [Authentication](https://platform.claude.com/docs/en/manage-claude/authentication).

236240

237241### April 30, 2026

238242

239* We've retired the 1M token context window beta (`context-1m-2025-08-07`) for Claude Sonnet 4.5 and Claude Sonnet 4. The beta header now has no effect on these models, and requests exceeding the standard 200k-token context window return an error. To use the 1M context window, migrate to [Claude Sonnet 4.6](https://platform.claude.com/docs/en/about-claude/models/overview#latest-models-comparison) or [Claude Opus 4.6](https://platform.claude.com/docs/en/about-claude/models/overview#latest-models-comparison), where it's generally available at standard pricing with no beta header required.

243* We've retired the 1M token context window beta (`context-1m-2025-08-07`) for Claude Sonnet 4.5 and Claude Sonnet 4. The beta header now has no effect on these models, and requests exceeding the standard 200k-token context window return an error. To use the 1M context window, migrate to [Claude Sonnet 4.6](https://platform.claude.com/docs/en/about-claude/models/overview#latest-models-comparison) or [Claude Opus 4.6](https://platform.claude.com/docs/en/about-claude/models/overview#latest-models-comparison), where it's included at standard pricing with no beta header required.

240244

241245### April 29, 2026

242246

from line 260

256260

257261### April 16, 2026

258262

259* We've launched [Claude Opus 4.7](https://www.anthropic.com/news/claude-opus-4-7), our most capable generally available model for complex reasoning and agentic coding, at the same $5 / $25 per MTok pricing as Opus 4.6. See [What's new in Claude Opus 4.7](https://platform.claude.com/docs/en/about-claude/models/whats-new-claude-4-7) for capability improvements, new features, and the updated tokenizer. Opus 4.7 includes API breaking changes versus Opus 4.6; see the [migration guide](https://platform.claude.com/docs/en/about-claude/models/migration-guide) before upgrading.

263* We've launched [Claude Opus 4.7](https://www.anthropic.com/news/claude-opus-4-7), our most capable widely released model for complex reasoning and agentic coding, at the same $5 / $25 per MTok pricing as Opus 4.6. See [What's new in Claude Opus 4.7](https://platform.claude.com/docs/en/about-claude/models/whats-new-claude-4-7) for capability improvements, new features, and the updated tokenizer. Opus 4.7 includes API breaking changes versus Opus 4.6; see the [migration guide](https://platform.claude.com/docs/en/about-claude/models/migration-guide) before upgrading.

260264* [Claude in Amazon Bedrock](https://platform.claude.com/docs/en/build-with-claude/claude-in-amazon-bedrock) is now open to all Amazon Bedrock customers. Claude Opus 4.7 and Claude Haiku 4.5 are available self-serve from the Bedrock console through the Messages API endpoint at `/anthropic/v1/messages`, in 27 AWS regions with global and regional endpoints.

261265* We've launched [task budgets](https://platform.claude.com/docs/en/build-with-claude/task-budgets) in beta on Claude Opus 4.7. Give Claude an advisory token budget for a full agentic loop (thinking, tool calls, tool results, and output) and the model sees a running countdown, using it to prioritize work and finish gracefully as the budget is consumed. Include the `task-budgets-2026-03-13` beta header in your requests.

262266* Claude Opus 4.7 supports [high-resolution image input](https://platform.claude.com/docs/en/build-with-claude/vision#high-resolution-image-support-on-claude-opus-4-7), raising the maximum image resolution from 1568 to 2576 pixels on the long edge for improved performance on computer use, screenshot understanding, and document analysis. High-resolution support is automatic and requires no beta header; images may use up to approximately 3x more image tokens than on prior models.

from line 299

295299

296300### March 13, 2026

297301

298* The [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) is now generally available for Claude Opus 4.6 and Sonnet 4.6 at standard pricing. Requests over 200k tokens work automatically for these models with no beta header required. The 1M token context window remains in beta for Claude Sonnet 4.5 and Sonnet 4.

302* The [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) is out of beta for Claude Opus 4.6 and Sonnet 4.6, at standard pricing. Requests over 200k tokens work automatically for these models with no beta header required. The 1M token context window remains in beta for Claude Sonnet 4.5 and Sonnet 4.

299303* We've removed the dedicated 1M rate limits for all supported models. Your standard account limits now apply across every context length.

300304* We've raised the media limit from 100 to 600 images or PDF pages per request when using the 1M token context window.

301305

from line 313

309313

310314* We've launched [Claude Sonnet 4.6](https://www.anthropic.com/news/claude-sonnet-4-6), our latest balanced model combining speed and intelligence for everyday tasks. Sonnet 4.6 delivers improved agentic search performance while consuming fewer tokens. Sonnet 4.6 supports [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) and a [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) (beta). See [Models & Pricing](https://platform.claude.com/docs/en/about-claude/models) for details.

311315* API [code execution](https://platform.claude.com/docs/en/agents-and-tools/tool-use/code-execution-tool) is now **free when used with web search or web fetch**. Sandboxed code execution improves model capability and token efficiency. See the [pricing details](https://platform.claude.com/docs/en/agents-and-tools/tool-use/code-execution-tool#usage-and-pricing) for standalone usage.

312* The [web search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool) and [programmatic tool calling](https://platform.claude.com/docs/en/agents-and-tools/tool-use/programmatic-tool-calling) are now generally available (no beta header required). Web search and web fetch now support [dynamic filtering](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool#dynamic-filtering), which uses code execution to filter results before they reach the context window for better performance and reduced token cost.

313* The [code execution tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/code-execution-tool), [web fetch tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-fetch-tool), [tool search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/tool-search-tool), [tool use examples](https://platform.claude.com/docs/en/agents-and-tools/tool-use/define-tools#providing-tool-use-examples), and [memory tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool) are now generally available (no beta header required).

316* The [web search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool) and [programmatic tool calling](https://platform.claude.com/docs/en/agents-and-tools/tool-use/programmatic-tool-calling) are available with no beta header required. Web search and web fetch now support [dynamic filtering](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool#dynamic-filtering), which uses code execution to filter results before they reach the context window for better performance and reduced token cost.

317* The [code execution tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/code-execution-tool), [web fetch tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-fetch-tool), [tool search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/tool-search-tool), [tool use examples](https://platform.claude.com/docs/en/agents-and-tools/tool-use/define-tools#providing-tool-use-examples), and [memory tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool) no longer require a beta header.

314318

315319### February 7, 2026

316320

from line 323

319323### February 5, 2026

320324

321325* We've launched [Claude Opus 4.6](https://www.anthropic.com/news/claude-opus-4-6), our most intelligent model for complex agentic tasks and long-horizon work. Opus 4.6 recommends [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/thinking) (`thinking: {type: "adaptive"}`); manual thinking (`type: "enabled"` with `budget_tokens`) is deprecated. Opus 4.6 does not support prefilling assistant messages. Learn more in [What's new in Claude 4.6](https://platform.claude.com/docs/en/about-claude/models/whats-new-claude-4-6).

322* The [effort parameter](https://platform.claude.com/docs/en/build-with-claude/effort) is now generally available (no beta header required) and supports Claude Opus 4.6. Effort replaces `budget_tokens` for controlling thinking depth on new models.

326* The [effort parameter](https://platform.claude.com/docs/en/build-with-claude/effort) no longer requires a beta header and now supports Claude Opus 4.6. Effort replaces `budget_tokens` for controlling thinking depth on new models.

323327* We've launched the [compaction API](https://platform.claude.com/docs/en/build-with-claude/compaction) in beta, providing server-side context summarization for effectively infinite conversations. Available on Opus 4.6.

324328* We've introduced [data residency controls](https://platform.claude.com/docs/en/manage-claude/data-residency), allowing you to specify where model inference runs with the `inference_geo` parameter. US-only inference is available at 1.1x pricing for models released after February 1, 2026.

325329* The [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) is now available in beta for Claude Opus 4.6, in addition to Sonnet 4.5 and Sonnet 4. [Long context pricing](https://platform.claude.com/docs/en/about-claude/pricing#long-context-pricing) applies to requests exceeding 200k input tokens.

326* [Fine-grained tool streaming](https://platform.claude.com/docs/en/agents-and-tools/tool-use/fine-grained-tool-streaming) is now generally available on all models and platforms (no beta header required).

330* [Fine-grained tool streaming](https://platform.claude.com/docs/en/agents-and-tools/tool-use/fine-grained-tool-streaming) no longer requires a beta header on any model or platform.

327331

328332### January 29, 2026

329333

330* [Structured outputs](https://platform.claude.com/docs/en/build-with-claude/structured-outputs) are now generally available on the Claude API for Claude Sonnet 4.5, Claude Opus 4.5, and Claude Haiku 4.5. GA includes expanded schema support, improved grammar compilation latency, and a simplified integration path with no beta header required. The `output_format` parameter has moved to `output_config.format`. Existing beta users can continue using the beta header during the transition period. Structured outputs remain in public beta on Amazon Bedrock and Microsoft Foundry.

334* [Structured outputs](https://platform.claude.com/docs/en/build-with-claude/structured-outputs) are out of beta on the Claude API for Claude Sonnet 4.5, Claude Opus 4.5, and Claude Haiku 4.5. This release includes expanded schema support, improved grammar compilation latency, and a simplified integration path with no beta header required. The `output_format` parameter has moved to `output_config.format`. Existing beta users can continue using the beta header during the transition period. Structured outputs remain in public beta on Amazon Bedrock and Microsoft Foundry.

331335

332336### January 12, 2026

333337

from line 359

355359

356360### November 21, 2025

357361

358* Search result content blocks are now generally available on Amazon Bedrock. Learn more in [Search results](https://platform.claude.com/docs/en/build-with-claude/search-results).

362* Search result content blocks are now available on Amazon Bedrock with no beta header required. Learn more in [Search results](https://platform.claude.com/docs/en/build-with-claude/search-results).

359363

360364### November 19, 2025

361365

from line 455

451455### August 13, 2025

452456

453457* We announced the deprecation of the Claude Sonnet 3.5 models (`claude-3-5-sonnet-20240620` and `claude-3-5-sonnet-20241022`). These models will be retired on October 28, 2025. We recommend migrating to Claude Sonnet 4.5 (`claude-sonnet-4-5-20250929`) for improved performance and capabilities. Read more in [Model deprecations](https://platform.claude.com/docs/en/about-claude/model-deprecations).

454* The 1-hour cache duration for prompt caching is now generally available. You can now use the extended cache TTL without a beta header. Learn more in [Prompt caching](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#1-hour-cache-duration).

458* The 1-hour cache duration for prompt caching no longer requires a beta header. Learn more in [Prompt caching](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#1-hour-cache-duration).

455459

456460### August 12, 2025

457461

from line 467

463467

464468### August 8, 2025

465469

466* Search result content blocks are now generally available on the Claude API and Vertex AI. This feature enables natural citations for RAG applications with proper source attribution. The beta header `search-results-2025-06-09` is no longer required. Learn more in [Search results](https://platform.claude.com/docs/en/build-with-claude/search-results).

470* Search result content blocks are out of beta on the Claude API and Vertex AI. This feature enables natural citations for RAG applications with proper source attribution. The beta header `search-results-2025-06-09` is no longer required. Learn more in [Search results](https://platform.claude.com/docs/en/build-with-claude/search-results).

467471

468472### August 5, 2025

469473

from line 516

512516* We've launched the [Code execution tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/code-execution-tool) in public beta, a tool that enables Claude to execute Python code in a secure, sandboxed environment.

513517* We've launched the [MCP connector](https://platform.claude.com/docs/en/agents-and-tools/mcp-connector) in public beta, a feature that allows you to connect to remote MCP servers directly from the Messages API.

514518* To increase answer quality and decrease tool errors, we've changed the default value for the `top_p` [nucleus sampling](https://en.wikipedia.org/wiki/Top-p_sampling) parameter in the Messages API from 0.999 to 0.99 for all models. To revert this change, set `top_p` to 0.999. Additionally, when extended thinking is enabled, you can now set `top_p` to values between 0.95 and 1.

515* We've moved our [Go SDK](https://github.com/anthropics/anthropic-sdk-go) from beta to GA.

519* Our [Go SDK](https://github.com/anthropics/anthropic-sdk-go) has moved from beta to its first stable release.

516520* We've included minute and hour level granularity to the [Usage](https://console.anthropic.com/settings/usage) page of Console alongside 429 error rates on the Usage page.

517521

518522### May 21, 2025

519523

520* We've moved our [Ruby SDK](https://github.com/anthropics/anthropic-sdk-ruby) from beta to GA.

524* Our [Ruby SDK](https://github.com/anthropics/anthropic-sdk-ruby) has moved from beta to its first stable release.

521525

522526### May 7, 2025

523527

from line 537

533537

534538### March 31st, 2025

535539

536* We've moved our [Java SDK](https://github.com/anthropics/anthropic-sdk-java) from beta to GA.

540* Our [Java SDK](https://github.com/anthropics/anthropic-sdk-java) has moved from beta to its first stable release.

537541* We've moved our [Go SDK](https://github.com/anthropics/anthropic-sdk-go) from alpha to beta.

538542

539543### February 27th, 2025

from line 594

590594

591595### December 17th, 2024

592596

593The following features are now generally available in the Claude API:

597The following features are now available in the Claude API without a beta header:

594598

595599* [Models API](https://platform.claude.com/docs/en/api/models/list): Query available models, validate model IDs, and resolve [model aliases](https://platform.claude.com/docs/en/about-claude/models/overview) to their canonical model IDs.

596600* [Message Batches API](https://platform.claude.com/docs/en/build-with-claude/batch-processing): Process large batches of messages asynchronously at 50% of the standard API cost.

from line 667

663667

664668### August 19th, 2024

665669

666* We've moved 8,192 token outputs from beta to general availability for Claude Sonnet 3.5.

670* 8,192-token outputs on Claude Sonnet 3.5 are out of beta and no longer require the `max-tokens-3-5-sonnet-2024-07-15` header.

667671

668672### August 14th, 2024

669673

from line 689

685689

686690### June 20th, 2024

687691

688* [Claude Sonnet 3.5](https://www.anthropic.com/news/claude-3-5-sonnet), our most intelligent model yet, is now generally available across the Claude API, Amazon Bedrock, and Vertex AI.

692* [Claude Sonnet 3.5](https://www.anthropic.com/news/claude-3-5-sonnet), our most intelligent model yet, is now available across the Claude API, Amazon Bedrock, and Vertex AI.

689693

690694### May 30th, 2024

691695

692* [Tool use](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview) is now generally available across the Claude API, Amazon Bedrock, and Vertex AI.

696* [Tool use](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview) is out of beta across the Claude API, Amazon Bedrock, and Vertex AI, with no beta header required.

693697

694698### May 10th, 2024

695699

about-claude/model-deprecations Changed · +1 / -1 lines

from line 213

213213

214214## API parameter deprecations

215215

216Anthropic occasionally deprecates request parameters that no longer apply to current models. Deprecated parameters remain in the SDK request types so existing code continues to type-check, but their behavior changes per model.

216Anthropic occasionally deprecates request parameters that no longer apply to current models. How the API treats a deprecated parameter depends on the model, as the following table shows. Most SDKs keep deprecated parameters in their request types so existing code continues to type-check. The Python SDK (v1.0 and later) removes `temperature`, `top_p`, and `top_k`, so passing them raises a `TypeError`.

217217

219219| ------------------------------- | -------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------- |

about-claude/models/introducing-claude-fable-5-and-claude-mythos-5 Changed · +2 / -2 lines

from line 50

5050

5151Claude Fable 5 and Claude Mythos 5 both become available on June 9, 2026:

5252

53* **Claude Fable 5** is generally available on the Claude API, [Amazon Bedrock](https://platform.claude.com/docs/en/build-with-claude/claude-in-amazon-bedrock), [Claude Platform on AWS](https://platform.claude.com/docs/en/build-with-claude/claude-platform-on-aws), [Google Cloud](https://platform.claude.com/docs/en/build-with-claude/claude-on-vertex-ai), and [Microsoft Foundry](https://platform.claude.com/docs/en/build-with-claude/claude-in-microsoft-foundry).

54* **Claude Mythos 5** is not generally available: it is offered in limited availability to approved customers in [Project Glasswing](https://anthropic.com/glasswing). For access, contact your Anthropic, AWS, or Google Cloud account team. Customers without access to Claude Mythos 5 can use Claude Fable 5, which is generally available and offers the same capabilities.

53* **Claude Fable 5** is available on the Claude API, [Amazon Bedrock](https://platform.claude.com/docs/en/build-with-claude/claude-in-amazon-bedrock), [Claude Platform on AWS](https://platform.claude.com/docs/en/build-with-claude/claude-platform-on-aws), [Google Cloud](https://platform.claude.com/docs/en/build-with-claude/claude-on-vertex-ai), and [Microsoft Foundry](https://platform.claude.com/docs/en/build-with-claude/claude-in-microsoft-foundry).

54* **Claude Mythos 5** is offered only to approved customers in [Project Glasswing](https://anthropic.com/glasswing). For access, contact your Anthropic, AWS, or Google Cloud account team. Customers without access to Claude Mythos 5 can use Claude Fable 5, which does not require access approval and offers the same capabilities.

5555

5656Claude Fable 5 and Claude Mythos 5 carry 30-day data retention and are not available under zero data retention: both are designated [Covered Models](https://support.claude.com/en/articles/15425695). See [Model-specific data retention requirements](https://platform.claude.com/docs/en/manage-claude/api-and-data-retention#model-specific-data-retention-requirements).

5757

about-claude/models/overview Changed · +1 / -1 lines

from line 16

1616

1717Claude Fable 5 (`claude-fable-5`) is Anthropic's most capable widely released model. Claude Mythos 5 (`claude-mythos-5`) shares Claude Fable 5's specs and pricing and joins the invitation-only Claude Mythos Preview (`claude-mythos-preview`) within [Project Glasswing](https://anthropic.com/glasswing). See [Introducing Claude Fable 5 and Claude Mythos 5](https://platform.claude.com/docs/en/about-claude/models/introducing-claude-fable-5-and-claude-mythos-5) for launch details and API changes.

1818

19Claude Fable 5 is generally available on the Claude API, Amazon Bedrock, Claude Platform on AWS, Google Cloud, and Microsoft Foundry beginning June 9, 2026. Claude Mythos 5 is not generally available: it is offered in limited availability to approved customers in [Project Glasswing](https://anthropic.com/glasswing), beginning the same day. For access, contact your Anthropic, AWS, or Google Cloud account team.

19Claude Fable 5 is available on the Claude API, Amazon Bedrock, Claude Platform on AWS, Google Cloud, and Microsoft Foundry beginning June 9, 2026. Claude Mythos 5 is offered only to approved customers in [Project Glasswing](https://anthropic.com/glasswing), beginning the same day. For access, contact your Anthropic, AWS, or Google Cloud account team.

2020

2121### Latest models comparison

2222

about-claude/models/whats-new-sonnet-5 Changed · +1 / -1 lines

from line 12

1212| --------------- | ----------------- | ---------------------------------------------- |

1313| Claude Sonnet 5 | `claude-sonnet-5` | The best combination of speed and intelligence |

1414

15Claude Sonnet 5 supports the [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) by default (1M tokens is both the default and the maximum; there is no smaller context variant), 128k max output tokens, [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/thinking), and the same set of tools and platform features as Claude Sonnet 4.6, except [Priority Tier](https://platform.claude.com/docs/en/api/service-tiers#supported-models), which is not available on Claude Sonnet 5. On the Claude API, Claude Sonnet 5 also supports the [browser use tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/browser-use-tool) and the generally available `computer_toolset_20260801` version of the [computer use tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool), neither of which Claude Sonnet 4.6 supports; the earlier `computer_20251124` version is still accepted on both models. To upgrade an existing integration, see [Migrate from `computer_20251124`](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool#migrate-from-computer-20251124).

15Claude Sonnet 5 supports the [1M token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows) by default (1M tokens is both the default and the maximum; there is no smaller context variant), 128k max output tokens, [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/thinking), and the same set of tools and platform features as Claude Sonnet 4.6, except [Priority Tier](https://platform.claude.com/docs/en/api/service-tiers#supported-models), which is not available on Claude Sonnet 5. On the Claude API, Claude Sonnet 5 also supports the [browser use tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/browser-use-tool) and the stable `computer_toolset_20260801` version of the [computer use tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool), neither of which Claude Sonnet 4.6 supports; the earlier `computer_20251124` version is still accepted on both models. To upgrade an existing integration, see [Migrate from `computer_20251124`](https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool#migrate-from-computer-20251124).

1616

1717For complete pricing and specs, see the [models overview](https://platform.claude.com/docs/en/about-claude/models/overview).

1818

about-claude/use-case-guides/customer-support-chat Changed · +1 / -1 lines

from line 9

99To follow this guide, you need:

1010

1111* A Claude API key (set as the `ANTHROPIC_API_KEY` environment variable)

12* Python 3.9 or later

12* Python 3.10 or later

1313

1414Install the required packages:

1515

about-claude/use-case-guides/ticket-routing Changed · +1 / -3 lines

from line 342

342342 message = client.messages.create(

343343 model=DEFAULT_MODEL,

344344 max_tokens=500,

345 temperature=0,

346345 messages=[{"role": "user", "content": classification_prompt}],

347346 stream=False,

348347 )

from line 408

409408 message = client.messages.create(

410409 model=DEFAULT_MODEL,

411410 max_tokens=500,

412 temperature=0,

413411 messages=[{"role": "user", "content": classification_prompt}],

414412 )

415413 usage = message.usage # Get the usage statistics for the API call for how many input and output tokens were used.

from line 459

461459

462460For example, you might have a top-level classifier that broadly categorizes tickets into "Technical Issues," "Billing Questions," and "General Inquiries." Each of these categories can then have its own sub-classifier to further refine the classification.

463461

464![](https://platform.claude.com/docs/images/ticket-hierarchy.png)

462![Classifier hierarchy routing tickets to Technical Issues, Billing Questions, or General Inquiries, each with a sub-classifier](https://platform.claude.com/docs/images/ticket-hierarchy.png)

465463

466464* **Pros - greater nuance and accuracy:** You can create different prompts for each parent path, allowing for more targeted and context-specific classification. This can lead to improved accuracy and more nuanced handling of customer requests.

467465

agents-and-tools/agent-skills/claude-api-skill Changed · +1 / -1 lines

from line 120

120120* **Cloud platform detection**, preserving platform-specific model ID formats (for example, the `anthropic.` prefix on Amazon Bedrock) and skipping changes for features that are unavailable on partner-operated platforms

121121* **Breaking parameter changes**, such as removing `temperature`, `top_p`, and `top_k` for Claude Opus 4.8 and Claude Opus 4.7, and converting `thinking: {type: "enabled", budget_tokens: N}` to `thinking: {type: "adaptive"}`

122122* **Prefill replacement**, converting assistant-message prefill patterns to [structured outputs](https://platform.claude.com/docs/en/build-with-claude/structured-outputs) where applicable

123* **Beta header cleanup**, removing headers that are GA on the target model (for example, `effort-2025-11-24`, `fine-grained-tool-streaming-2025-05-14`, `interleaved-thinking-2025-05-14`) and switching back from `client.beta.messages.create` to `client.messages.create`

123* **Beta header cleanup**, removing beta headers that the target model doesn't require (for example, `effort-2025-11-24`, `fine-grained-tool-streaming-2025-05-14`, `interleaved-thinking-2025-05-14`) and switching back from `client.beta.messages.create` to `client.messages.create`

124124* **Effort calibration**, recommending an `output_config.effort` starting point for the target model (for example, the default `high` on Claude Opus 5, and `xhigh` for coding and agentic use cases on Claude Opus 4.8 and Claude Opus 4.7)

125125* **Prompt-behavior tuning**, flagging length-control, tool-triggering, subagent, and instruction-following prompts that may behave differently on the target model

126126* **Silent default handling**, opting back into thinking summarization (`thinking.display: "summarized"`) when reasoning is surfaced to users on Claude Opus 4.8 and Claude Opus 4.7

agents-and-tools/mcp-connector Changed · +1 / -2 lines

from line 1317

13171317 ```

13181318

13191319 ```php PHP

1320 // The PHP SDK exposes the Files API under the beta namespace; field names can differ from other SDKs.

13211320 // As a content block in a message

13221321 $resource = $mcp->readResource('file:///path/to/doc.txt');

13231322

from line 1338

13391338

13401339 // As a file upload

13411340 $fileResource = $mcp->readResource('file:///path/to/data.json');

1342 $file = $anthropic->beta->files->upload(file: BetaMcp::resourceToFile($fileResource));

1341 $file = $anthropic->files->upload(file: BetaMcp::resourceToFile($fileResource));

13431342 echo $file->id, "\n";

13441343 ```

13451344

agents-and-tools/tool-use/memory-tool Changed · +2 / -2 lines

from line 104

104104

105105## Getting started

106106

107The memory tool is generally available on the Messages API: no beta header is required. Using it takes two steps:

107Using the memory tool takes two steps:

108108

1091091. Add the memory tool to your request. The `tools` entry `{"type": "memory_20250818", "name": "memory"}` is the entire configuration: the `name` must be `memory`, and you don't define an input schema for an Anthropic-provided tool.

1101102. Implement a client-side handler for each memory command. Your handler must reject paths outside `/memories`, so read [Path traversal protection](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool#path-traversal-protection) before you write it.

from line 285

285285

286286Claude's reply to a request like the previous one ends with a `tool_use` block that requests a memory operation, such as `view /memories`. Your application executes the operation and returns the result in a `tool_result` block, then sends the conversation back so Claude can continue: the standard [tool-use loop](https://platform.claude.com/docs/en/agents-and-tools/tool-use/handle-tool-calls).

287287

288Four SDKs provide memory tool helpers that handle the tool interface and the loop. Subclass `BetaAbstractMemoryTool` (Python and C#), use `betaMemoryTool` (TypeScript), or implement `BetaMemoryToolHandler` (Java) to back memory with your own storage, such as files on disk, a database, cloud storage, or encrypted files. Python and TypeScript also ship a ready-made local-filesystem implementation, `BetaLocalFilesystemMemoryTool`. The helper and tool-runner surfaces live in each SDK's beta namespace even though the memory tool itself is generally available. The Go and Ruby SDKs have no memory helper, so those examples run the tool-use loop themselves, and PHP wraps your handler closure in its generic `BetaRunnableTool`. All three use an in-memory store that you replace with your own storage.

288Four SDKs provide memory tool helpers that handle the tool interface and the loop. Subclass `BetaAbstractMemoryTool` (Python and C#), use `betaMemoryTool` (TypeScript), or implement `BetaMemoryToolHandler` (Java) to back memory with your own storage, such as files on disk, a database, cloud storage, or encrypted files. Python and TypeScript also ship a ready-made local-filesystem implementation, `BetaLocalFilesystemMemoryTool`. The helper and tool-runner surfaces live in each SDK's beta namespace even though the memory tool itself doesn't require a beta header. The Go and Ruby SDKs have no memory helper, so those examples run the tool-use loop themselves, and PHP wraps your handler closure in its generic `BetaRunnableTool`. All three use an in-memory store that you replace with your own storage.

289289

290290<CodeGroup exclude="shell">

291291 ```python Python

agents-and-tools/tool-use/tool-runner Changed · +1 / -1 lines

from line 1088

10881088

10891089### Automatic context management

10901090

1091For long-running agentic tasks, the Python, TypeScript, and Ruby tool runners support automatic [compaction](https://platform.claude.com/docs/en/build-with-claude/context-editing#client-side-compaction-sdk), which generates summaries when token usage exceeds a threshold so the conversation can continue beyond context window limits. All three SDKs have deprecated this client-side option in favor of server-side [context editing](https://platform.claude.com/docs/en/build-with-claude/context-editing), which is available in every SDK. The Go, Java, C#, and PHP tool runners don't include client-side compaction.

1091For long-running agentic tasks, the TypeScript and Ruby tool runners support automatic [compaction](https://platform.claude.com/docs/en/build-with-claude/context-editing#client-side-compaction-sdk), which generates summaries when token usage exceeds a threshold so the conversation can continue beyond context window limits. Both SDKs have deprecated this client-side option in favor of [server-side compaction](https://platform.claude.com/docs/en/build-with-claude/compaction), which works with every SDK's tool runner through the `context_management` request parameter. The Python SDK (v1.0 and later) and the Go, Java, C#, and PHP tool runners don't include client-side compaction.

10921092

10931093### Debugging tool execution

10941094

api/admin/analytics/artifacts Changed · +1 / -1 lines

from line 100

100100```http

101101curl https://api.anthropic.com/v1/organizations/analytics/artifacts \

102102 -H 'anthropic-version: 2023-06-01' \

103 -H "Authorization: Bearer $ANTHROPIC_OAUTH_TOKEN"

103 -H "X-Api-Key: $ANTHROPIC_ADMIN_API_KEY"

104104```

105105

106106#### Response

api/admin/analytics/artifacts/list Changed · +1 / -1 lines

from line 98

9898```http

9999curl https://api.anthropic.com/v1/organizations/analytics/artifacts \

100100 -H 'anthropic-version: 2023-06-01' \

101 -H "Authorization: Bearer $ANTHROPIC_OAUTH_TOKEN"

101 -H "X-Api-Key: $ANTHROPIC_ADMIN_API_KEY"

102102```

103103

104104#### Response

api/admin/analytics/chat_projects/list Changed · +1 / -1 lines

from line 136

136136```http

137137curl https://api.anthropic.com/v1/organizations/analytics/apps/chat/projects \

138138 -H 'anthropic-version: 2023-06-01' \

139 -H "Authorization: Bearer $ANTHROPIC_OAUTH_TOKEN"

139 -H "X-Api-Key: $ANTHROPIC_ADMIN_API_KEY"

140140```

141141

142142#### Response

One read of Claude Developer Platformapi-20260821T200735Z

What this read moved

build-with-claude/vision-coordinates Changed · +31 / -1 lines

claude_api_primer Changed · +3 / -4 lines

cli-sdks-libraries/sdks/python Changed · +20 / -18 lines

manage-claude/compliance-activity-feed Changed · +4 / -2 lines

manage-claude/compliance-integration-patterns Changed · +10 / -7 lines

managed-agents/cloud-sandboxes-reference Changed · +37 / -25 lines

managed-agents/define-outcomes Changed · +7 / -11 lines

managed-agents/environments Changed · +5 / -7 lines

managed-agents/files Changed · +15 / -14 lines

managed-agents/github Changed · +10 / -4 lines

managed-agents/session-operations Changed · +5 / -5 lines

release-notes/overview Changed · +29 / -25 lines

about-claude/model-deprecations Changed · +1 / -1 lines

about-claude/models/introducing-claude-fable-5-and-claude-mythos-5 Changed · +2 / -2 lines

about-claude/models/overview Changed · +1 / -1 lines

about-claude/models/whats-new-sonnet-5 Changed · +1 / -1 lines

about-claude/use-case-guides/customer-support-chat Changed · +1 / -1 lines

about-claude/use-case-guides/ticket-routing Changed · +1 / -3 lines

agents-and-tools/agent-skills/claude-api-skill Changed · +1 / -1 lines

agents-and-tools/mcp-connector Changed · +1 / -2 lines

agents-and-tools/tool-use/memory-tool Changed · +2 / -2 lines

agents-and-tools/tool-use/tool-runner Changed · +1 / -1 lines

api/admin/analytics/artifacts Changed · +1 / -1 lines

api/admin/analytics/artifacts/list Changed · +1 / -1 lines

api/admin/analytics/chat_projects/list Changed · +1 / -1 lines