api-20260922T163722Z / documentation capture

build-with-claude/context-windows Changed · +1 / -1 lines

from line 33

3333

3434## Context window sizes by model

3535

36Claude Fable 5.1, Claude Mythos 5.1, Claude Fable 5, Claude Mythos 5, Claude Opus 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, Claude Sonnet 5, Claude Sonnet 4.6, and [Claude Mythos Preview](https://anthropic.com/glasswing) have a 1M-token context window. A single request to any of them can generate up to 128k output tokens (`max_tokens`). Other Claude models, including Claude Sonnet 4.5, have a 200k-token context window.

36Claude Fable 5.1, Claude Mythos 5.1, Claude Fable 5, Claude Mythos 5, Claude Opus 5.5, Claude Opus 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, Claude Sonnet 5, Claude Sonnet 4.6, and [Claude Mythos Preview](https://anthropic.com/glasswing) have a 1M-token context window. A single request to any of them can generate up to 128k output tokens (`max_tokens`). Other Claude models, including Claude Sonnet 4.5, have a 200k-token context window.

3737

3838For every model with a 1M-token context window, 1M is the default: you don't need a beta header, and long-context requests are billed at [standard pricing](https://platform.claude.com/docs/en/about-claude/pricing#long-context-pricing).

3939

build-with-claude/extended-thinking Changed · +1 / -1 lines

from line 354

354354You need to migrate off `type: "enabled"` if:

355355

356356* You use Claude Opus 4.6 or Claude Sonnet 4.6, where `budget_tokens` is deprecated.

357* You are moving to Claude Opus 4.7, Claude Opus 4.8, Claude Opus 5, Claude Sonnet 5, Claude Fable 5.1, Claude Mythos 5.1, Claude Fable 5, or Claude Mythos 5, where `type: "enabled"` returns a 400 error.

357* You are moving to Claude Opus 4.7, Claude Opus 4.8, Claude Opus 5, Claude Opus 5.5, Claude Sonnet 5, Claude Fable 5.1, Claude Mythos 5.1, Claude Fable 5, or Claude Mythos 5, where `type: "enabled"` returns a 400 error.

358358

359359The mapping is small: remove `budget_tokens`, set `thinking: {type: "adaptive"}`, and control reasoning depth with `output_config: {effort: ...}` instead of a token budget.

360360

build-with-claude/fallback-credit Changed · +1 / -1 lines

from line 622

622622 * **`fallback-credit-*`:** keep this header on both requests. The retry needs it to redeem the token.

623623

624624 <Note>

625 On models that include the 1M token context window by default, such as Claude Fable 5.1, Claude Fable 5, Claude Opus 5, and Claude Opus 4.8, the `context-1m-2025-08-07` beta header has no effect. To keep the two requests identical, omit that header on both rather than sending it on one and not the other.

625 On models that include the 1M token context window by default, such as Claude Fable 5.1, Claude Fable 5, Claude Opus 5.5, Claude Opus 5, and Claude Opus 4.8, the `context-1m-2025-08-07` beta header has no effect. To keep the two requests identical, omit that header on both rather than sending it on one and not the other.

626626 </Note>

627627</Accordion>

628628

build-with-claude/prompt-engineering/prompting-claude-fable-5-1 Changed · +1 / -1 lines

from line 753

753753

754754## Keep the conversation history append-only

755755

756Append each assistant turn to the history exactly as the API returned it, thinking blocks included, and don't edit earlier turns between requests. For new accounts created on or after August 31, 2026, Claude Fable 5.1's thinking blocks are valid [only in the exact conversation that produced them](https://platform.claude.com/docs/en/build-with-claude/thinking#preserved-in-conversation): a request that replays a thinking block after its prefix (the system prompt, the tool list, or any earlier message) has changed returns a 400, or drops the affected blocks if you set `thinking.block_binding.prefix_mismatch_behavior: "drop_block"` (beta, `thinking-binding-controls-2026-08-01` header). Future models are expected to enforce this check for all accounts, so adopt the pattern now even if yours isn't enforced today.

756Append each assistant turn to the history exactly as the API returned it, thinking blocks included, and don't edit earlier turns between requests. For new accounts created on or after August 31, 2026, Claude Fable 5.1's thinking blocks are valid [only in the exact conversation that produced them](https://platform.claude.com/docs/en/build-with-claude/thinking#preserved-in-conversation): a request that replays a thinking block after its prefix (the system prompt, the tool list, or any earlier message) has changed returns a 400, or drops the affected blocks if you set `thinking.block_binding.prefix_mismatch_behavior: "drop_block"` (beta, `thinking-binding-controls-2026-08-01` header). Adopt the pattern even if the check isn't enforced for your account, so the same code works on every account.

757757

758758The history edits that trip the check are the same ones that restart the [prompt cache](https://platform.claude.com/docs/en/build-with-claude/prompt-caching): injecting and removing per-turn reminders, summarizing older turns in place, or changing the system prompt mid-session. Send per-turn reminders as [turn-scoped system messages](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages#turn-scoped-system-messages), change instructions or tools with a [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages) instead of rewriting `system` or `tools`, and let server-side [compaction](https://platform.claude.com/docs/en/build-with-claude/compaction) or [context editing](https://platform.claude.com/docs/en/build-with-claude/context-editing) do any trimming. If you compact on the client, the simplest shape is to replace the whole history with one summary message plus the new user turn and replay nothing else: no thinking blocks carry over, so nothing fails, and the model thinks afresh on the compacted conversation (see [Custom compaction on the client](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#custom-compaction-on-the-client)). Because cache reads are now cheaper (see [Pricing](https://platform.claude.com/docs/en/models/fable-5-1/whats-new-fable-5-1#pricing)), compacting early to save cost may no longer be the right cost-intelligence tradeoff on Claude Fable 5.1, so experiment with later compaction points.

759759

build-with-claude/refusals-and-fallback Changed · +1 / -1 lines

from line 4

44description: How Claude Fable and Claude Opus models return classifier refusals and how to retry refused requests on a fallback model.

55---

66

7Claude Fable 5.1, Claude Fable 5, and Claude Opus 5 include safety classifiers that can decline a request. When that happens, you receive a normal response, not an error, with `stop_reason: "refusal"`. Its `stop_details.category` names the policy area (see [What a refusal looks like](https://platform.claude.com/docs/en/build-with-claude/refusals-and-fallback#refusal-response)). You can usually still get an answer by sending the same request to another Claude model. This page shows you how to recognize a refusal and how to set up that retry.

7Claude Fable 5.1, Claude Fable 5, Claude Opus 5.5, and Claude Opus 5 include safety classifiers that can decline a request. When that happens, you receive a normal response, not an error, with `stop_reason: "refusal"`. Its `stop_details.category` names the policy area (see [What a refusal looks like](https://platform.claude.com/docs/en/build-with-claude/refusals-and-fallback#refusal-response)). You can usually still get an answer by sending the same request to another Claude model. This page shows you how to recognize a refusal and how to set up that retry.

88

99Read this page when you build on any of these models and want declined requests to fall through to another model automatically. It also applies when you have seen `"refusal"` in a response and want to know what to do next.

1010

cli-sdks-libraries/cli/authentication Changed · +1 / -1 lines

from line 78

7878```bash CLI

7979ant messages create \

8080 --workspace-id wrkspc_01... \

81 --model claude-opus-5 \

81 --model claude-opus-5-5 \

8282 --max-tokens 1024 \

8383 --message '{role: user, content: "Hello, Claude"}'

8484```

cli-sdks-libraries/cli/scripting Changed · +1 / -1 lines

from line 118

118118 -H "anthropic-version: 2023-06-01" \

119119 -H "content-type: application/json" \

120120 -d '{

121 "model": "claude-opus-5",

121 "model": "claude-opus-5-5",

122122 "max_tokens": 256,

123123 "messages": [{"role": "user", "content": "hi"}]

124124 }'

cli-sdks-libraries/libraries/apple-foundation-models Changed · +2 / -2 lines

from line 68

6868Model identifiers are values of `ClaudeModel`. Use a compiled-in constant, or construct one with explicit capabilities for an ID that isn't compiled in yet (see [Capabilities](https://platform.claude.com/docs/en/cli-sdks-libraries/libraries/apple-foundation-models#capabilities)):

6969

7070```swift

71ClaudeLanguageModel(name: .opus5, auth: auth)

71ClaudeLanguageModel(name: .opus5_5, auth: auth)

7272```

7373

7474Constants mirror API model IDs (`.opus5` is `claude-opus-5`) and carry each model's capabilities. New models ship as new constants in package releases; check `ClaudeModel` in Xcode for the current list, and the [Models overview](https://platform.claude.com/docs/en/models/overview) to compare models.

from line 90

9090Pin a Claude [effort level](https://platform.claude.com/docs/en/build-with-claude/effort) for every request with `fixedEffort:`. It takes precedence over the framework's per-request reasoning hints. The framework's named reasoning levels stop at high; to request more effort for a single request instead, pass a custom reasoning level naming the Claude effort (`.custom("xhigh")` or `.custom("max")`), which maps directly. The API defaults to `high` when no effort is sent:

9191

9292```swift

93ClaudeLanguageModel(name: .opus5, auth: auth, fixedEffort: .xhigh)

93ClaudeLanguageModel(name: .opus5_5, auth: auth, fixedEffort: .xhigh)

9494```

9595

9696The level must be one the model accepts. Each `ClaudeModel` declares which of the five levels (`low`, `medium`, `high`, `xhigh`, `max`) its model takes, if any: some models don't accept effort at all.

home Changed · +1 / -1 lines

from line 242

242242 <HomeSectionHeader label="Models" title="The Claude model family" description="Choose the right model for your use case." />

243243

244244 * [Claude Fable 5.1](https://platform.claude.com/docs/en/models/fable-5-1/overview) (`claude-fable-5-1`) — New — *For demanding reasoning and long-horizon agentic work* — Most capable · Research · Multi-day tasks

245 * [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) (`claude-opus-5`) — *For complex agentic coding and enterprise work* — Complex projects · Agents · Coding

245 * [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) (`claude-opus-5-5`) — New — *For long-running agentic coding and knowledge work* — Complex projects · Agents · Coding

246246 * [Claude Sonnet 5](https://platform.claude.com/docs/en/models/sonnet-5/overview) (`claude-sonnet-5`) — *The best combination of speed and intelligence* — Everyday tasks · Writing · Cost-efficient

247247 * [Claude Haiku 4.5](https://platform.claude.com/docs/en/models/haiku-4-5/overview) (`claude-haiku-4-5`) — *The fastest model with near-frontier intelligence* — Fastest · Lowest cost · High volume

248248 </HomeSection>

intro Changed · +2 / -2 lines

from line 19

1919

2020## Explore the latest generation of Claude models

2121

22If you're unsure which model to use, start with [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) for most workloads. Use [Claude Fable 5.1](https://platform.claude.com/docs/en/models/fable-5-1/overview) for demanding reasoning and long-horizon agentic work, or when your evals on Claude Opus 5 at higher effort still fall short. All current models support text and image input, text output, multilingual capabilities, vision, and tool use. Each model's page lists the platforms it's available on.

22If you're unsure which model to use, start with [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) for most workloads. Use [Claude Fable 5.1](https://platform.claude.com/docs/en/models/fable-5-1/overview) for demanding reasoning and long-horizon agentic work, or when your evals on Claude Opus 5.5 at higher effort still fall short. All current models support text and image input, text output, multilingual capabilities, vision, and tool use. Each model's page lists the platforms it's available on.

2323

2424* [Claude Fable 5.1](https://platform.claude.com/docs/en/models/fable-5-1/overview) (`claude-fable-5-1`) — New — *For demanding reasoning and long-horizon agentic work* — Most capable · Research · Multi-day tasks

25* [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) (`claude-opus-5`) — *For complex agentic coding and enterprise work* — Complex projects · Agents · Coding

25* [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) (`claude-opus-5-5`) — New — *For long-running agentic coding and knowledge work* — Complex projects · Agents · Coding

2626* [Claude Sonnet 5](https://platform.claude.com/docs/en/models/sonnet-5/overview) (`claude-sonnet-5`) — *The best combination of speed and intelligence* — Everyday tasks · Writing · Cost-efficient

2727* [Claude Haiku 4.5](https://platform.claude.com/docs/en/models/haiku-4-5/overview) (`claude-haiku-4-5`) — *The fastest model with near-frontier intelligence* — Fastest · Lowest cost · High volume

2828

manage-claude/claude-code-analytics-api Changed · +2 / -2 lines

from line 188

188188 },

189189 "model_breakdown": [

190190 {

191 "model": "claude-opus-5",

191 "model": "claude-opus-5-5",

192192 "tokens": {

193193 "input": 100000,

194194 "output": 35000,

from line 197

197197 },

198198 "estimated_cost": {

199199 "currency": "USD",

200 "amount": 141

200 "amount": 113

201201 }

202202 }

203203 ]

manage-claude/cmek-azure-key-vault Changed · +1 / -1 lines

from line 157

157157 </Step>

158158

159159 <Step title="Attach the key to a workspace">

160 Open **Settings > Workspaces**, choose the workspace, and open its **Security** tab. Under **Encryption key**, select the key, click **Save**, and confirm. Attaching a key can't be undone. For a workspace that already receives requests, the key can take [up to a day to take effect](https://platform.claude.com/docs/en/manage-claude/cmek#how-it-works).

160 In the Claude Console, go to [Manage > Security](https://platform.claude.com/settings/workspaces/default/security-compliance) and select the workspace in the workspace picker at the top of the sidebar. Under **Encryption key**, select the key, click **Save**, and confirm. Attaching a key can't be undone. For a workspace that already receives requests, the key can take [up to a day to take effect](https://platform.claude.com/docs/en/manage-claude/cmek#how-it-works).

161161 </Step>

162162 </Steps>

163163 </Tab>

manage-claude/cmek-google-cloud-kms Changed · +1 / -1 lines

from line 150

150150 </Step>

151151

152152 <Step title="Attach the key to a workspace">

153 Open **Settings > Workspaces**, choose the workspace, and open its **Security** tab. Under **Encryption key**, select the key, click **Save**, and confirm. Attaching a key can't be undone. For a workspace that already receives requests, the key can take [up to a day to take effect](https://platform.claude.com/docs/en/manage-claude/cmek#how-it-works).

153 In the Claude Console, go to [Manage > Security](https://platform.claude.com/settings/workspaces/default/security-compliance) and select the workspace in the workspace picker at the top of the sidebar. Under **Encryption key**, select the key, click **Save**, and confirm. Attaching a key can't be undone. For a workspace that already receives requests, the key can take [up to a day to take effect](https://platform.claude.com/docs/en/manage-claude/cmek#how-it-works).

154154 </Step>

155155 </Steps>

156156 </Tab>

manage-claude/compliance-content-data Changed · +2 / -2 lines

from line 46

4646 "updated_at": "2026-04-10T09:10:11Z",

4747 "deleted_at": null,

4848 "href": "https://claude.ai/chat/abcdef01-2345-6789-abcd-ef0123456789",

49 "model": "claude-opus-5",

49 "model": "claude-opus-5-5",

5050 "organization_uuid": "91012d09-e48b-438e-a489-1bebfd8fa6f9",

5151 "project_id": "claude_proj_01KGp4eZNug9ri4kE35RSppq",

5252 "user": {

from line 100

100100 "updated_at": "2026-04-10T09:10:11Z",

101101 "deleted_at": null,

102102 "href": "https://claude.ai/chat/abcdef01-2345-6789-abcd-ef0123456789",

103 "model": "claude-opus-5",

103 "model": "claude-opus-5-5",

104104 "organization_uuid": "91012d09-e48b-438e-a489-1bebfd8fa6f9",

105105 "project_id": "claude_proj_01KGp4eZNug9ri4kE35RSppq",

106106 "user": {

manage-claude/compliance-sessions Changed · +2 / -2 lines

from line 201

201201 "type": "compliance_local_session_message",

202202 "id": "clsm_01J4KpLmNoPqRsTuVwXyZaBd",

203203 "role": "assistant",

204 "model": "claude-opus-5",

204 "model": "claude-opus-5-5",

205205 "created_at": "2026-07-09T14:02:11Z",

206206 "provenance": null,

207207 "content": [

from line 246

246246 "type": "compliance_local_session_message",

247247 "id": "clsm_01J4KpLmNoPqRsTuVwXyZaBf",

248248 "role": "assistant",

249 "model": "claude-opus-5",

249 "model": "claude-opus-5-5",

250250 "created_at": "2026-07-09T14:02:38Z",

251251 "provenance": null,

252252 "content": [

manage-claude/rate-limits-api Changed · +2 / -2 lines

from line 285

285285 {

286286 "type": "rate_limit",

287287 "group_type": "model_group",

288 "models": ["claude-opus-5"],

288 "models": ["claude-opus-5-5"],

289289 "limits": [

290290 { "type": "requests_per_minute", "value": 4000 },

291291 { "type": "input_tokens_per_minute", "value": 10000000 },

from line 617

617617 {

618618 "type": "workspace_rate_limit",

619619 "group_type": "model_group",

620 "models": ["claude-opus-5"],

620 "models": ["claude-opus-5-5"],

621621 "limits": [

622622 { "type": "requests_per_minute", "value": 1000, "org_limit": 4000 },

623623 { "type": "input_tokens_per_minute", "value": 500000, "org_limit": 10000000 }

manage-claude/usage-cost-api Changed · +1 / -1 lines

from line 125

125125curl "https://api.anthropic.com/v1/organizations/usage_report/messages?\

126126starting_at=2025-01-15T00:00:00Z&\

127127ending_at=2025-01-15T23:59:59Z&\

128models[]=claude-opus-5&\

128models[]=claude-opus-5-5&\

129129service_tiers[]=batch&\

130130context_window[]=0-200k&\

131131bucket_width=1h" \

managed-agents/environments Changed · +3 / -0 lines

from line 41

4141

4242 <File filename="environment.yaml">

4343 ```yaml

44 # yaml-language-server: $schema=https://platform.claude.com/schemas/ant/beta/environment.json

4445 name: python-dev

4546 config:

4647 type: cloud

from line 257

256257

257258 <File filename="environment.yaml">

258259 ```yaml

260 # yaml-language-server: $schema=https://platform.claude.com/schemas/ant/beta/environment.json

259261 name: data-analysis

260262 config:

261263 type: cloud

from line 436

434436

435437 <File filename="environment.yaml">

436438 ```yaml

439 # yaml-language-server: $schema=https://platform.claude.com/schemas/ant/beta/environment.json

437440 name: api-access

438441 config:

439442 type: cloud

managed-agents/events-and-streaming Changed · +1 / -1 lines

from line 2637

26372637### Sending system messages

26382638

26392639<Note>

2640 `system.message` is supported by Claude Fable 5.1, Claude Mythos 5.1, Claude Fable 5, Claude Mythos 5, Claude Opus 5, and Claude Opus 4.8. If the agent's primary model does not support mid-conversation system injection, the event is rejected with a `model_does_not_support_mid_conversation_system` validation error. Subagent models are not checked, because `system.message` lands on the primary thread only.

2640 `system.message` is supported by Claude Fable 5.1, Claude Mythos 5.1, Claude Fable 5, Claude Mythos 5, Claude Opus 5.5, Claude Opus 5, and Claude Opus 4.8. If the agent's primary model does not support mid-conversation system injection, the event is rejected with a `model_does_not_support_mid_conversation_system` validation error. Subagent models are not checked, because `system.message` lands on the primary thread only.

26412641</Note>

26422642

26432643Send a `system.message` event to give the agent privileged system-level context that applies to the accompanying turn and all subsequent turns. Unlike the `system` field on the agent definition (which sets the top-level system prompt), `system.message` content is appended to the session's system context as a `role: "system"` turn rather than replacing that prompt. Use it when the agent needs updated system-level guidance mid-session: a different persona, revised constraints, or context fetched at runtime that should shape the model's behavior going forward.

managed-agents/memory Changed · +1 / -0 lines

from line 44

4444

4545 <File filename="memory_store.yaml">

4646 ```yaml

47 # yaml-language-server: $schema=https://platform.claude.com/schemas/ant/beta/memory_store.json

4748 name: User Preferences

4849 description: Per-user preferences and project context.

4950 ```

release-notes/system-prompts/overview Changed · +2 / -0 lines

from line 7

77Claude's web interface ([claude.ai](https://claude.ai)) and mobile apps use a system prompt to provide up-to-date information, such as the current date, to Claude at the start of every conversation. The system prompt also encourages certain behaviors, such as always providing code snippets in Markdown. This prompt is periodically updated to improve Claude's responses. These system prompt updates do not apply to the Claude API. Some models have multiple dated entries on their pages. Starting with the Claude 4.6 generation, each model ID is a [single fixed snapshot](https://platform.claude.com/docs/en/about-claude/models/model-ids-and-versions), so those models have one entry.

88

99<CardGroup cols={3}>

10 <Card id="claude-opus-5-5" title="Claude Opus 5.5" icon="file" href="https://platform.claude.com/docs/en/release-notes/system-prompts/claude-opus-5-5" />

11

1012 <Card id="claude-fable-5-1" title="Claude Fable 5.1" icon="file" href="https://platform.claude.com/docs/en/release-notes/system-prompts/claude-fable-5-1" />

1113

1214 <Card id="claude-opus-5" title="Claude Opus 5" icon="file" href="https://platform.claude.com/docs/en/release-notes/system-prompts/claude-opus-5" />

resources/overview Changed · +4 / -0 lines

from line 5