HomeDocs changes
Documentation changes
What the docs quietly changed.
Anthropic's published documentation, fetched on a schedule and diffed against the last copy. Every edit dated, including the ones that never made a release note.
Last poll read all 4 cleanly: Claude Code CLI, Claude Developer Platform, Claude Documentation, Model Context Protocol. A date here is when this site noticed, not when Anthropic edited.
Claude Developer Platform
821pages tracked
767 changed this week · last read 24 Oct 2026
Claude Code CLI
228pages tracked
145 changed this week · last read 24 Oct 2026
Claude Documentation
280pages tracked
65 changed this week · last read 23 Oct 2026
Model Context Protocol
357pages tracked
171 changed this week · last read 23 Oct 2026
100 on this page
ReadSourcePageWhatLines
23 Sep 2026
333 pages
+97834 −37961 lines
claude-docs
third-party/claude-desktop/entra-broker
±changed
+11−11
claude-docschanged
third-party/claude-desktop/entra-broker
| Feature | How to enable it | Page
| ------------------------------------------------------------ | -------------------------------------------------------- | --------------------------------------------------------
| Microsoft Foundry inference provider | Set `inferenceFoundryAuthFlow` to `broker` | [Microsoft Foundry](/docs/third-party/claude-desktop/fou
| LLM gateway single sign-on | Set `inferenceGatewayOidcAuthFlow` to `broker` | [LLM gateway](/docs/third-party/claude-desktop/gateway#s
| Workforce Identity sign-in for Google Cloud's Agent Platform | Set `inferenceVertexWorkforceAuthFlow` to `broker` | [Google Cloud's Agent Platform](/docs/third-party/claude
+11 −11 lines
Upstream edited 23 Sep 2026 05:32 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-docs
third-party/claude-desktop/ssh-remote-sessions
±changed
+1−1
claude-docschanged
third-party/claude-desktop/ssh-remote-sessions
| [Amazon Bedrock](/docs/third-party/claude-desktop/bedrock) | Bearer token, credential helper | In-app AWS sign-in (IAM Identity Cente
| [Amazon Bedrock](/docs/third-party/claude-desktop/bedrock) | Bearer token, identity provider sign-in, credential helper | In-app AWS sign-in (IAM Identity Cente
+1 −1 lines
Upstream edited 23 Sep 2026 05:32 UTC Recorded here 23 Sep 2026 05:37 UTC
claude-docs
third-party/claude-desktop/gateway
±changed
+2−0
claude-docschanged
third-party/claude-desktop/gateway
From Claude Desktop 2.7032.0, `inferenceCredentialKind: "external-idp"` with [`inferenceIdpOidc`](/docs/third-party/claude-desktop/configuration#inferenceidpoidc) and `inferenceIdp
+2 −0 lines
Upstream edited 23 Sep 2026 05:32 UTC Recorded here 23 Sep 2026 05:37 UTC
claude-docs
third-party/claude-desktop/entra-broker
±changed
+3−2
claude-docschanged
third-party/claude-desktop/entra-broker
| Amazon Bedrock identity provider sign-in (through a proxy) | Set `inferenceIdpAuthFlow` to `broker` | [Amazon Bedrock](/docs/third-party/claude-desktop/bedroc
For the gateway and Workforce Identity flows, the broker is available only when your identity provider is Microsoft Entra ID: the `issuer` in `inferenceGatewayOidc` or `inferenceVe
For the gateway, Amazon Bedrock, and Workforce Identity flows, the broker is available only when your identity provider is Microsoft Entra ID: the `issuer` in `inferenceGatewayOidc
What happens on Linux, or on a Windows or macOS device where the broker is unavailable, depends on the feature. For the inference sign-in flows (Foundry, gateway, and Workforce Ide
What happens on Linux, or on a Windows or macOS device where the broker is unavailable, depends on the feature. For the inference sign-in flows (Foundry, gateway, Amazon Bedrock, a
+3 −2 lines
Upstream edited 23 Sep 2026 05:32 UTC Recorded here 23 Sep 2026 05:37 UTC
claude-docs
third-party/claude-desktop/bedrock
±changed
+46−15
claude-docschanged
third-party/claude-desktop/bedrock
| Scenario | Use | Per-device prerequisite | Per-use
| ------------------------------------------ | --------------------------------------------------------------------------------- | --------------------------------------- | -------
| Proof of concept, single team | [Bearer token](#bearer-token) (`inferenceBedrockBearerToken`) | None | No (sha
| Broad rollout to users without AWS tooling | [In-app AWS sign-in](#in-app-aws-sign-in) (`inferenceBedrockSso*`) | None | Yes
| Developers who already use the AWS CLI | [Named profile](#named-profile) (`inferenceBedrockProfile`) | AWS CLI v2 and a pushed `~/.aws/config` | Yes
+46 −15 lines
Upstream edited 23 Sep 2026 05:32 UTC Recorded here 23 Sep 2026 05:37 UTC
claude-docs
office-agents/performance
±changed
+2−2
claude-docschanged
office-agents/performance
moves to the next. See [Work across M365 apps](/docs/office-agents/work-across-apps.md).
moves to the next. See [Work across M365 apps](https://claude.com/docs/office-agents/work-across-apps.md).
[Configure a custom OpenTelemetry collector](/docs/office-agents/opentelemetry.md).
[Configure a custom OpenTelemetry collector](https://claude.com/docs/office-agents/opentelemetry.md).
+2 −2 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 23 Oct 2026 18:07 UTC
claude-docs
office-agents/performance
±changed
+2−2
claude-docschanged
office-agents/performance
moves to the next. See [Work across M365 apps](/docs/office-agents/work-across-apps).
moves to the next. See [Work across M365 apps](/docs/office-agents/work-across-apps.md).
[Configure a custom OpenTelemetry collector](/docs/office-agents/opentelemetry).
[Configure a custom OpenTelemetry collector](/docs/office-agents/opentelemetry.md).
+2 −2 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 22 Oct 2026 18:07 UTC
claude-docs
office-agents/performance
±changed
+109−113
claude-docschanged
office-agents/performance
responding.<sup id="cite-ref-1-a" className="scroll-mt-24"><a href="#cite-note-1">\[1]</a></sup> This happens with any heavy task in Office, with
responding.<sup id="cite-ref-1-a">[\[1\]](#cite-note-1)</sup> This happens with any heavy task in Office, with
Office is busy, other actions can stall too.<sup id="cite-ref-1-b" className="scroll-mt-24"><a href="#cite-note-1">\[1]</a></sup>
<img src="https://mintcdn.com/claude-ai/_Xjykq_jOPUQ7vjH/images/office-agents/performance/perf-one-step.png?fit=max&auto=format&n=_Xjykq_jOPUQ7vjH&q=85&s=86e807e77450672873ddcbd4e1
+109 −113 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 20 Oct 2026 18:07 UTC
claude-docs
office-agents/third-party-platforms
±changed
+112−112
claude-docschanged
office-agents/third-party-platforms
| Path | How it works |
| ---------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------- |
| LLM gateway | Requests route through your gateway (LiteLLM, Portkey, Kong, and others) to your chosen provider. Matches the pattern used by Claude Code. |
| Bedrock direct | The add-in authenticates via Microsoft Entra ID and calls Amazon Bedrock directly without intermediaries. |
| Vertex AI direct | The add-in authenticates through Google OAuth and calls Vertex AI directly. |
+112 −112 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-docs
office-agents/performance
±changed
+30−30
claude-docschanged
office-agents/performance
| Limit | Value | Wh
| ------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------- | --
| Wait for a step where Claude runs code | 90 seconds | A
| Wait for other steps, such as writing cells | 5 minutes on the desktop, 2 minutes on the web | A
| Wait for a Word text edit | 30 seconds | A
+30 −30 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-docs
office-agents/opentelemetry
±changed
+74−74
claude-docschanged
office-agents/opentelemetry
| Setting | Description |
| --------------- | --------------------------------------------------------------------------------- |
| `otlp_endpoint` | Base URL of your OTLP collector. The add-in appends `/v1/traces` |
| `otlp_headers` | Optional authentication headers in OpenTelemetry `key1=value1,key2=value2` format |
| Setting | Description |
+74 −74 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 28 Sep 2026 22:07 UTC
claude-docs
office-agents/performance
+new page
+312−0
claude-docsnew page
office-agents/performance
No line of the body moved in this read.
+312 −0 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 23 Sep 2026 13:37 UTC
claude-docs
office-agents/third-party-platforms
±changed
+85−27
claude-docschanged
office-agents/third-party-platforms
| Domain | Required when | Purpose |
| ---------------------------------------- | ------------------------- | ------------------------------------------------------------------------------------- |
| `pivot.claude.ai` | Always | Add-in host serving task pane UI, analytics, and telemetry. |
| `claude.ai/api/` | Always | Feature-flag evaluation without sign-in. |
| `appsforoffice.microsoft.com` | Always | Microsoft Office.js runtime script. |
+85 −27 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 23 Sep 2026 03:07 UTC
claude-docs
office-agents/opentelemetry
±changed
+2−2
claude-docschanged
office-agents/opentelemetry
custom manifest.
custom manifest, with `otlp_headers` after `#` so the credential stays out of [server request logs](/docs/office-agents/third-party-platforms#keep-credentials-out-of-the-manifest-u
https://pivot.claude.ai/taskpane.html?otlp_endpoint=https://otel-collector.example.com&otlp_headers=Authorization=Bearer%20<token>
https://pivot.claude.ai/taskpane.html?otlp_endpoint=https://otel-collector.example.com#otlp_headers=Authorization=Bearer%20<token>
+2 −2 lines
Upstream edited 23 Sep 2026 02:40 UTC Recorded here 23 Sep 2026 03:07 UTC
claude-code
vs-code
±changed
+22−1
claude-codechanged
vs-code
You can also type `/plan` in the prompt box. Requires Claude Code v2.1.280 or later.
* `/plan`: switches to plan mode. If you're already in plan mode, shows the current plan instead.
* `/plan` with a task, such as `/plan fix the auth bug`: switches to plan mode and starts planning that task.
+22 −1 lines
Upstream edited 23 Sep 2026 02:03 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-docs
claude-science/core-concepts
±changed
+1−1
claude-docschanged
claude-science/core-concepts
| Run code | Run Python code? / Run a shell command? / Install packages? | Once or Always |
| Run code | Run Python code? / Run a shell command? / Install packages? | Once, This conversation, This project, or Global |
+1 −1 lines
Upstream edited 23 Sep 2026 01:57 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-docs
claude-science/changelog
±changed
+10−0
claude-docschanged
claude-science/changelog
<Update label="0.1.52" description="September 22, 2026">
* Clicking a figure in Claude's answer now opens it beside the chat; Ctrl/Cmd-click or the new "Open fullscreen" button opens it full screen
* Opening or closing the right pane now keeps your place in a long answer
* Markdown files no longer rotate in Safari after they load
+10 −0 lines
Upstream edited 23 Sep 2026 01:57 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-docs
claude-science/artifacts
±changed
+1−1
claude-docschanged
claude-science/artifacts
Click a linked file in the conversation to open it in a tab beside the chat. HTML artifacts have zoom controls, including fit to width; images zoom up to their native resolution. O
Click a linked file or a figure in the conversation to open it in a tab beside the chat. Ctrl/Cmd-click opens it full screen. HTML artifacts have zoom controls, including fit to wi
+1 −1 lines
Upstream edited 23 Sep 2026 01:57 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
voice-dictation
±changed
+10−0
claude-codechanged
voice-dictation
Holding or tapping `Space` starts dictation only where the keypress would otherwise type into the prompt. In the [transcript viewer](/docs/en/interactive-mode#transcript-viewer), `
## Cancel a recording
+10 −0 lines
Upstream edited 23 Sep 2026 01:51 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
plugin-marketplaces
±changed
+12−28
claude-codechanged
plugin-marketplaces
By default, the background refresh disables git credential helpers when it checks the marketplace's remote for new commits, so the check can't authenticate to private repositories
When the check finds new commits, or fails because it can't reach or authenticate to the remote, Claude Code clones the marketplace again and swaps the new clone in. If that clone
The background refresh checks the marketplace's remote for new commits with your configured git credential helpers, the same as the commands you run. For SSH remotes, a key loaded
+12 −28 lines
Upstream edited 23 Sep 2026 01:49 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
monitoring-usage
±changed
+5−0
claude-codechanged
monitoring-usage
* `stdout_chars`: Total characters of stdout across the matching hooks that succeeded. Requires Claude Code v2.1.280 or later
* `additional_context_chars`: Total characters of `additionalContext` returned by the matching hooks. Requires Claude Code v2.1.280 or later
* `system_message_chars`: Total characters of `systemMessage` returned by the matching hooks. Requires Claude Code v2.1.280 or later
* `initial_user_message_chars`: Total characters of `initialUserMessage` returned by the matching hooks. Requires Claude Code v2.1.280 or later
* `num_outputs_persisted`: Number of hook outputs over the [10,000-character cap](/docs/en/hooks#json-output) that Claude Code saved to a file. Requires Claude Code v2.1.280 or lat
+5 −0 lines
Upstream edited 23 Sep 2026 01:49 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
artifacts
±changed
+4−0
claude-codechanged
artifacts
Whether artifacts are allowed for your organization comes from your organization's policy, which Claude Code loads from `api.anthropic.com`. When Claude Code can't load the policy,
If a proxy, VPN, or web filter is involved, ask your IT admin to let `api.anthropic.com` through. Claude Code keeps retrying in the background, and artifacts become available once
+4 −0 lines
Upstream edited 23 Sep 2026 01:49 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
skills
±changed
+8−0
claude-codechanged
skills
### Personal skills disappeared
If skill folders you created in `~/.claude/skills/` are gone, look in `~/.claude/skills/.trash/`. When Claude Code [syncs skills from claude.ai](#how-synced-skills-behave), it down
Before v2.1.280, a file named `manifest.json` in `~/.claude/skills/` caused Claude Code to move the skill folders that file listed into a timestamped folder under `~/.claude/skills
+8 −0 lines
Upstream edited 23 Sep 2026 01:49 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
ultrareview
±changed
+19−19
claude-codechanged
ultrareview
| Plan | Included free runs | After free runs |
| ------------------- | ------------------ | ------------------------------------------------------------------------------------------------------------ |
| Pro | 3 free runs | billed as [usage credits](https://support.claude.com/en/articles/12429409-extra-usage-for-paid-claude-plans) |
| Max | 3 free runs | billed as [usage credits](https://support.claude.com/en/articles/12429409-extra-usage-for-paid-claude-plans) |
| Team and Enterprise | none | billed as [usage credits](https://support.claude.com/en/articles/12429409-extra-usage-for-paid-claude-plans) |
+19 −19 lines
Upstream edited 23 Sep 2026 01:48 UTC Recorded here 28 Sep 2026 23:37 UTC
claude-code
ultrareview
±changed
+8−2
claude-codechanged
ultrareview
Claude can also tell you that a review was stopped or that its session wasn't found:
* If the review's cloud session is stopped or [archived](/docs/en/claude-code-on-the-web#archive-sessions) on claude.ai before the review finishes, Claude tells you it was stopped.
* If the review's cloud session was deleted, or you've signed in to a different Claude account or organization since launching it, Claude tells you the session wasn't found.
+8 −2 lines
Upstream edited 23 Sep 2026 01:48 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
plugins
±changed
+2−0
claude-codechanged
plugins
To load plugins in a session where you can't add the flag, list their absolute paths in the [`CLAUDE_CODE_PLUGIN_DIRS`](/docs/en/env-vars#variables) environment variable instead. C
+2 −0 lines
Upstream edited 23 Sep 2026 01:47 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
mcp
±changed
+3−1
claude-codechanged
mcp
Claude Code truncates tool descriptions and server instructions at 2KB each. Keep them concise to avoid truncation, and put critical details near the start.
Claude Code truncates each tool description and each server's instructions at 2,048 characters by default. Keep them concise, and put critical details near the start.
To change the limit for every MCP server in your session, set [`CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH`](/docs/en/env-vars#variables) to a number of characters. This variable requi
+3 −1 lines
Upstream edited 23 Sep 2026 01:47 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-code
env-vars
±changed
+2−0
claude-codechanged
env-vars
| `CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH` | Maximum length in characters of each MCP tool description and each MCP server's instructions that Claude Code sends to t
| `CLAUDE_CODE_PLUGIN_DIRS` | Plugin directories to load for the session, each loaded the way a [`--plugin-dir`](/docs/en/plugins#test-your-plugins-lo
+2 −0 lines
Upstream edited 23 Sep 2026 01:47 UTC Recorded here 23 Sep 2026 02:07 UTC
claude-docs
third-party/claude-desktop/telemetry
±changed
+1−0
claude-docschanged
third-party/claude-desktop/telemetry
| `claude.ai` | Performance timing |
+1 −0 lines
Upstream edited 23 Sep 2026 00:34 UTC Recorded here 23 Sep 2026 00:37 UTC
claude-docs
third-party/claude-desktop/gateway
±changed
+52−17
claude-docschanged
third-party/claude-desktop/gateway
| <span id="inferencegatewayoidcauthflow" />Gateway sign-in flow<br />`inferenceGatewayOidcAuthFlow` | `enum` | MDM + Bootstrap<br />Added in 1.25927.0 | — | How the IdP
| <span id="inferencegatewayoidc" />Gateway SSO IdP (OIDC)<br />`inferenceGatewayOidc` | `object` | MDM + Bootstrap<br />Added in 1.6889.0 | — | External Id
| <span id="inferencegatewayoidcauthflow" />Gateway sign-in flow<br />`inferenceGatewayOidcAuthFlow` | `enum` | MDM + Bootstrap<br />Added in 1.25927.0 | — | How the IdP
| <span id="inferencegatewayoidc" />Gateway SSO IdP (OIDC)<br />`inferenceGatewayOidc` | `object` | MDM + Bootstrap<br />Added in 1.6889.0 | — | External Id
| <span id="inferenceidpauthflow" />Identity provider sign-in flow<br />`inferenceIdpAuthFlow` | `enum` | MDM + Bootstrap<br />Added in 2.7032.0 | — | How the ide
+52 −17 lines
Upstream edited 23 Sep 2026 00:34 UTC Recorded here 23 Sep 2026 00:37 UTC
claude-docs
third-party/claude-desktop/configuration-changelog
±changed
+56−0
claude-docschanged
third-party/claude-desktop/configuration-changelog
<Update label="v2.7032.0" description="2026-09-22">
<div className="cfg-keys">
| MDM key | Type | Description |
| ------------------------------------------------------------------------------------------------------------------------ | --------- | ------------------------------------- |
+56 −0 lines
Upstream edited 23 Sep 2026 00:34 UTC Recorded here 23 Sep 2026 00:37 UTC
claude-docs
third-party/claude-desktop/configuration
±changed
+118−56
claude-docschanged
third-party/claude-desktop/configuration
| Setting | Type | Availability
| ------------------------------------------------------------------------------------------------------------------------------------------------ | ---------- | ------------------
| <span id="inferencecustomheaders" />Custom inference headers<br />`inferenceCustomHeaders` | `object` | MDM + Bootstrap<br
| <span id="inferencesessionlifetimesec" />Sign-in session lifetime<br />`inferenceSessionLifetimeSec` | `integer` | MDM + Bootstrap<br
| <span id="inferencecredentialhelper" />Helper script<br />`inferenceCredentialHelper` | `string` | MDM + Bootstrap<br
+118 −56 lines
Upstream edited 23 Sep 2026 00:34 UTC Recorded here 23 Sep 2026 00:37 UTC
claude-docs
cowork/changelog
±changed
+46−0
claude-docschanged
cowork/changelog
<Update label="v2.7032.0" description="2026-09-22">
Bundled Claude Code version: 2.1.280.
**General**
+46 −0 lines
Upstream edited 23 Sep 2026 00:34 UTC Recorded here 23 Sep 2026 00:37 UTC
claude-docs
claude-tag/admins/attach-to-scope
±changed
+8−7
claude-docschanged
claude-tag/admins/attach-to-scope
Three kinds of standing instruction can apply in a channel, written by different people:
The table lists the kinds of standing instruction that can apply in a channel and who writes each.
| Layer | Who writes it
| :------------------ | :-----------------------------------------------------------------------------------------------------------------------------------------------------------
| Custom instructions | Owner for any scope; channel members and [channel managers](/docs/claude-tag/admins/restrict-access#delegate-channel-setup-to-channel-managers) for the chan
+8 −7 lines
Upstream edited 23 Sep 2026 00:02 UTC Recorded here 23 Sep 2026 00:07 UTC
22 Sep 2026
274 pages
+8048 −4478 lines
claude-code
quickstart
±changed
+5−5
claude-codechanged
quickstart
```bash theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null}
```bash theme={null}
```powershell theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null}
```powershell theme={null}
```batch theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null}
+5 −5 lines
Upstream edited 22 Sep 2026 23:25 UTC Recorded here 22 Sep 2026 23:37 UTC
claude-code
overview
±changed
+5−5
claude-codechanged
overview
```bash theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null}
```bash theme={null}
```powershell theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null}
```powershell theme={null}
```batch theme={null} theme={null} theme={null} theme={null} theme={null} theme={null} theme={null}
+5 −5 lines
Upstream edited 22 Sep 2026 23:25 UTC Recorded here 22 Sep 2026 23:37 UTC
claude-docs
claude-tag/admins/restrict-access
±changed
+8−4
claude-docschanged
claude-tag/admins/restrict-access
**Channel name rules** steer where Claude works by channel name instead of channel by channel. The rules sit in the **Advanced** section of the **Default Slack access** panel and o
**Channel name rules** steer where Claude works by channel name instead of channel by channel. The rules sit in the **Advanced** section of the **Default Slack access** panel and o
* **Auto-join channel patterns**: Claude joins a public channel whose name matches when the channel is created or renamed. Private channels still need an invite. To add Claude to a
A pattern is written in lowercase, like Slack channel names, plus two wildcards: `*` matches any run of characters and `?` matches exactly one. `inc-*` matches every channel whose
+8 −4 lines
Upstream edited 22 Sep 2026 23:24 UTC Recorded here 22 Sep 2026 23:37 UTC
claude-docs
claude-tag/admins/attach-to-scope
±changed
+19−1
claude-docschanged
claude-tag/admins/attach-to-scope
Attaching binds the bundle to a workspace scope or to a single channel under it.
Attaching binds the bundle to a workspace scope, to a single channel under it, or to every channel whose name matches a pattern.
### Attach a bundle to channels by name
+19 −1 lines
Upstream edited 22 Sep 2026 23:24 UTC Recorded here 22 Sep 2026 23:37 UTC
claude-code
prompt-caching
±changed
+6−6
claude-codechanged
prompt-caching
* **Effort level**: on most models, each effort level has its own cache, so changing effort mid-session recomputes the entire request. On Fable 5.1 with an API key or a Claude subs
* **Effort level**: on most models, each effort level has its own cache, so changing effort mid-session recomputes the entire request. On Opus 5.5 and Fable 5.1 with an API key or
On Fable 5.1 with an API key or a Claude subscription, changing effort keeps the cache, and Claude Code applies the new level without asking. This doesn't apply on Amazon Bedrock,
On Opus 5.5 and Fable 5.1 with an API key or a Claude subscription, changing effort keeps the cache, and Claude Code applies the new level without asking. This doesn't apply on Ama
| Field | Meaning |
+6 −6 lines
Upstream edited 22 Sep 2026 23:19 UTC Recorded here 22 Sep 2026 23:37 UTC
claude-docs
claude-tag/concepts/personal-connectors
±changed
+10−4
claude-docschanged
claude-tag/concepts/personal-connectors
In a channel, an admin decides the shared access. The channel uses the connections an admin attached to it, and everyone who asks there gets the same access. In organizations where
In organizations where connector use in channels is available, Claude can use your personal connectors for the requests you make in a channel.
In a direct message (DM), your connectors apply on their own, because a DM runs on [your own claude.ai account](/docs/claude-tag/concepts/agent-identity#direct-message-channels).
* **Approval**: Claude asks you before it starts using your connectors in a channel.
* **Access**: Claude works with your permissions and reaches only what your account can reach.
+10 −4 lines
Upstream edited 22 Sep 2026 23:03 UTC Recorded here 22 Sep 2026 23:07 UTC
claude-docs
claude-tag/admins/federated-access/gcp
±changed
+1−1
claude-docschanged
claude-tag/admins/federated-access/gcp
If Claude reports HTTP 403 with a reason that starts with [`request blocked: federated connections work only in agent sessions (such as a Slack channel), not in personal sessions (
If Claude reports HTTP 403 with a reason that starts with [`request blocked: federated connections work only in agent sessions (such as a Slack channel), not in personal sessions (
+1 −1 lines
Upstream edited 22 Sep 2026 22:47 UTC Recorded here 22 Sep 2026 23:07 UTC
claude-docs
claude-tag/admins/federated-access/connect-a-gateway
±changed
+1−1
claude-docschanged
claude-tag/admins/federated-access/connect-a-gateway
If Claude reports HTTP 403 with a reason that starts with [`request blocked: federated connections work only in agent sessions (such as a Slack channel), not in personal sessions (
If Claude reports HTTP 403 with a reason that starts with [`request blocked: federated connections work only in agent sessions (such as a Slack channel), not in personal sessions (
+1 −1 lines
Upstream edited 22 Sep 2026 22:47 UTC Recorded here 22 Sep 2026 23:07 UTC
claude-docs
claude-tag/admins/federated-access/aws
±changed
+1−1
claude-docschanged
claude-tag/admins/federated-access/aws
If Claude reports HTTP 403 with a reason that starts with [`request blocked: federated connections work only in agent sessions (such as a Slack channel), not in personal sessions (
If Claude reports HTTP 403 with a reason that starts with [`request blocked: federated connections work only in agent sessions (such as a Slack channel), not in personal sessions (
+1 −1 lines
Upstream edited 22 Sep 2026 22:47 UTC Recorded here 22 Sep 2026 23:07 UTC
claude-docs
claude-tag/admins/federated-access/authorization-server
±changed
+1−1
claude-docschanged
claude-tag/admins/federated-access/authorization-server
If Claude reports HTTP 403 with a reason that starts with [`request blocked: federated connections work only in agent sessions (such as a Slack channel), not in personal sessions (
If Claude reports HTTP 403 with a reason that starts with [`request blocked: federated connections work only in agent sessions (such as a Slack channel), not in personal sessions (
+1 −1 lines
Upstream edited 22 Sep 2026 22:47 UTC Recorded here 22 Sep 2026 23:07 UTC
claude-docs
claude-tag/admins/restrict-access
±changed
+2−0
claude-docschanged
claude-tag/admins/restrict-access
A channel manager is a person. To let the members of another Slack channel write a channel's instructions, see [Manage a channel's instructions from another channel](/docs/claude-t
+2 −0 lines
Upstream edited 22 Sep 2026 22:05 UTC Recorded here 22 Sep 2026 22:07 UTC
claude-docs
claude-tag/admins/managed-by
+new page
+165−0
claude-docsnew page
claude-tag/admins/managed-by
No line of the body moved in this read.
+165 −0 lines
Upstream edited 22 Sep 2026 22:05 UTC Recorded here 22 Sep 2026 22:07 UTC
claude-docs
claude-tag/admins/attach-to-scope
±changed
+2−0
claude-docschanged
claude-tag/admins/attach-to-scope
To let a central team write a channel's instructions from its own Slack channel, see [Manage a channel's instructions from another channel](/docs/claude-tag/admins/managed-by).
+2 −0 lines
Upstream edited 22 Sep 2026 22:05 UTC Recorded here 22 Sep 2026 22:07 UTC
claude-code
output-styles
±changed
+108−36
claude-codechanged
output-styles
> Adapt Claude Code for uses beyond software engineering
Output styles change how Claude responds, not what Claude knows. They set Claude's role, tone, and output format for every response. Use one when you keep re-prompting for the same
A custom output style gives Claude your own instructions and lets you choose whether to keep Claude Code's built-in software engineering instructions. Keep them when you're changin
+108 −36 lines
Upstream edited 22 Sep 2026 20:16 UTC Recorded here 22 Sep 2026 20:37 UTC
claude-code
features-overview
±changed
+39−19
claude-codechanged
features-overview
* **[Output styles](/docs/en/output-styles)** set Claude's role, tone, and response format for every response in a session
| Feature | What it does | When to use it
| -------------------------------------------------------------- | ---------------------------------------------------------------------------- | ----------------------------------
| **CLAUDE.md** | Persistent context loaded every conversation | Project conventions, "always do X"
| **Skill** | Instructions, knowledge, and workflows Claude can use | Reusable content, reference docs,
+39 −19 lines
Upstream edited 22 Sep 2026 19:59 UTC Recorded here 22 Sep 2026 20:07 UTC
claude-code
llm-gateway-protocol
±changed
+1−0
claude-codechanged
llm-gateway-protocol
* When the gateway or its upstream rejects the [advisor tool](/docs/en/advisor) entry in `tools` as an unrecognized tool type, Claude Code retries the request once without that ent
+1 −0 lines
Upstream edited 22 Sep 2026 17:28 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
advisor
±changed
+1−1
claude-codechanged
advisor
* **Anthropic API only**: the advisor is a server-executed tool. It is not available on Amazon Bedrock, Claude Platform on AWS, Google Cloud's Agent Platform, or Microsoft Foundry.
* **Anthropic API only**: the advisor is a server-executed tool. It is not available on Amazon Bedrock, Claude Platform on AWS, Google Cloud's Agent Platform, or Microsoft Foundry.
+1 −1 lines
Upstream edited 22 Sep 2026 17:28 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
sub-agents
±changed
+10−6
claude-codechanged
sub-agents
The `--agents` flag accepts JSON with a `prompt` field plus these [frontmatter](#supported-frontmatter-fields) fields: `description`, `tools`, `disallowedTools`, `model`, `permissi
The `--agents` flag accepts JSON with a `prompt` field plus these [frontmatter](#supported-frontmatter-fields) fields: `description`, `tools`, `disallowedTools`, `model`, `permissi
#### Supported frontmatter fields
The following fields can be used in the YAML frontmatter. Only `name` and `description` are required.
+10 −6 lines
Upstream edited 22 Sep 2026 17:25 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
skills
±changed
+4−4
claude-codechanged
skills
Beyond the markdown content, you can configure skill behavior using YAML frontmatter fields between `---` markers at the top of your `SKILL.md` file:
Configure a skill with YAML [frontmatter](/docs/en/glossary#frontmatter) between `---` markers at the top of `SKILL.md`, and write the skill's instructions as Markdown after the cl
All fields are optional. Only `description` is recommended so Claude knows when to use the skill.
Claude Code reads the frontmatter only when the opening `---` is the file's first line. Otherwise it treats the whole file, `---` markers included, as skill content.
+4 −4 lines
Upstream edited 22 Sep 2026 17:25 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
plugins-reference
±changed
+53−14
claude-codechanged
plugins-reference
Plugin agents support `name`, `description`, `model`, `effort`, `maxTurns`, `tools`, `disallowedTools`, `skills`, `memory`, `background`, [`omitClaudeMd`](/docs/en/sub-agents#suppo
For security reasons, plugin-shipped agents don't support `hooks`, `mcpServers`, or `permissionMode`.
#### Plugin agent frontmatter
+53 −14 lines
Upstream edited 22 Sep 2026 17:25 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
output-styles
±changed
+10−8
claude-codechanged
output-styles
### Frontmatter
<h3 id="frontmatter">
Frontmatter reference
</h3>
Output style files support these frontmatter fields:
+10 −8 lines
Upstream edited 22 Sep 2026 17:25 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
memory
±changed
+12−0
claude-codechanged
memory
<h4 id="rules-frontmatter-reference">
Rule frontmatter reference
</h4>
+12 −0 lines
Upstream edited 22 Sep 2026 17:25 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
glossary
±changed
+8−0
claude-codechanged
glossary
## F
### Frontmatter
+8 −0 lines
Upstream edited 22 Sep 2026 17:25 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
claude-directory
±changed
+14−0
claude-codechanged
claude-directory
## Frontmatter fields by file
Skills, command files, subagents, output styles, and rules read their configuration from YAML [frontmatter](/docs/en/glossary#frontmatter) at the top of the file, and each accepts
| File | Frontmatter fields
+14 −0 lines
Upstream edited 22 Sep 2026 17:25 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
discover-plugins
±changed
+1−1
claude-codechanged
discover-plugins
* The marketplace wasn't [added from GitHub, another Git host, or a remote URL](#add-marketplaces).
* The marketplace wasn't [added from GitHub, another Git host, a remote URL](#add-marketplaces), or [claude.ai](#add-from-claude-ai).
+1 −1 lines
Upstream edited 22 Sep 2026 17:24 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
settings-reference
±changed
+2−3
claude-codechanged
settings-reference
Have Claude respond in a language other than English by default. There is no fixed list for responses: Claude Code adds the value verbatim to the system prompt as an instruction to
Have Claude respond in a language other than English by default. There is no fixed list for responses: Claude Code passes the value verbatim to Claude as an instruction to always r
Claude Code enforces these lists at the OS sandbox boundary, so they apply to every subprocess a sandboxed command starts, such as `kubectl`, `terraform`, or `npm`, not only to Cla
Claude Code enforces these lists at the OS sandbox boundary, so they apply to every subprocess a sandboxed command starts, such as `kubectl`, `terraform`, or `npm`. Claude Code add
| `npm` | `{ "source": "npm", "package": "@acme-corp/claude-plugins" }` | `package` required, the npm pack
+2 −3 lines
Upstream edited 22 Sep 2026 17:24 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
network-config
±changed
+1−0
claude-codechanged
network-config
| `github.com` | Cloning GitHub-hosted [plugin marketplaces](/docs/en/plugin-marketplaces) and plugins, including the official Anthropic marketplace, over H
+1 −0 lines
Upstream edited 22 Sep 2026 17:23 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
errors
±changed
+2−1
claude-codechanged
errors
* On macOS and Linux, an entry path that contains a backslash anywhere after the leading `./`
* If you maintain the marketplace, write the entry's `source` as a plain relative path such as `./plugins/my-plugin`, and keep any symlink it crosses pointed inside the marketplace
* If you maintain the marketplace, write the entry's `source` as a plain relative path with forward slashes, such as `./plugins/my-plugin`, and keep any symlink it crosses pointed
+2 −1 lines
Upstream edited 22 Sep 2026 17:22 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-docs
cowork/changelog
±changed
+1−0
claude-docschanged
cowork/changelog
* Added support for Claude Opus 5.5.
+1 −0 lines
Upstream edited 22 Sep 2026 17:15 UTC Recorded here 22 Sep 2026 17:37 UTC
mcp
community/working-groups/infrastructure
±changed
+1−1
mcpchanged
community/working-groups/infrastructure
The [Working and Interest Groups governance rules](/community/working-interest-groups) apply.
The [Working and Interest Groups governance rules](/community/working-interest-groups.md) apply.
+1 −1 lines
Upstream edited 22 Sep 2026 17:08 UTC Recorded here 22 Oct 2026 20:37 UTC
mcp
community/working-groups/infrastructure
±changed
+17−17
mcpchanged
community/working-groups/infrastructure
| Role | Name | Organization | GitHub | Term |
| ---- | ----------------------- | ------------ | ---------------------------------------------------------------------- | ------- |
| Lead | David Soria Parra (DSP) | Anthropic | [@dsp](https://github.com/dsp), [@dsp-ant](https://github.com/dsp-ant) | Initial |
| Lead | Den Delimarsky | Anthropic | [@localden](https://github.com/localden) | Initial |
| Lead | Sambhav Kothari | Bloomberg | [@sambhav](https://github.com/sambhav) | Initial |
+17 −17 lines
Upstream edited 22 Sep 2026 17:08 UTC Recorded here 28 Sep 2026 22:07 UTC
mcp
community/working-groups/infrastructure
+new page
+111−0
mcpnew page
community/working-groups/infrastructure
No line of the body moved in this read.
+111 −0 lines
Upstream edited 22 Sep 2026 17:08 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
agent-sdk/typescript
±changed
+42−7
claude-codechanged
agent-sdk/typescript
readMcpResource(serverName: string, uri: string): Promise<SDKControlMcpReadResourceResponse>;
| `mcpServerStatus()` | Returns status of connected MCP servers
| `mcpServerStatus()` | Returns the status of connected MCP servers as [`McpServerStatus`](#mcpserverstatus)`[]`
| `readMcpResource(serverName, uri)` | *Alpha.* Reads one MCP Apps `ui://` resource from a connected MCP server so your application can render a tool's widget. Resolves with an
### `SDKControlMcpReadResourceResponse`
+42 −7 lines
Upstream edited 22 Sep 2026 17:07 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
keybindings
±changed
+1−1
claude-codechanged
keybindings
| `chat:clearInput` | Ctrl+L | Force a full screen redraw, preserving input and conversation. In [fullscreen rendering](/docs/en/fullscreen#clear-the
| `chat:clearInput` | Ctrl+L | Force a full screen redraw, preserving input and conversation
+1 −1 lines
Upstream edited 22 Sep 2026 17:07 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
interactive-mode
±changed
+1−1
claude-codechanged
interactive-mode
| `Ctrl+L` | Redraw or clear the screen
| `Ctrl+L` | Redraw the screen
+1 −1 lines
Upstream edited 22 Sep 2026 17:07 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
fullscreen
±changed
+3−3
claude-codechanged
fullscreen
To clear the screen and keep the conversation, press `Ctrl+L`. The earlier messages scroll up out of view, and you can scroll back with `PgUp` or the mouse wheel to read them again
`Cmd+K` does the same as `Ctrl+L` when your terminal passes it through to Claude Code. iTerm2 and Terminal.app handle `Cmd+K` themselves, and Claude Code redraws the conversation i
If the display looks garbled or partially blank, press `Ctrl+L` to redraw the screen. The redraw keeps the conversation and your input in place.
+3 −3 lines
Upstream edited 22 Sep 2026 17:07 UTC Recorded here 22 Sep 2026 17:37 UTC
claude-code
remote-control
±changed
+1−1
claude-codechanged
remote-control
* **Diff of your changes**: when the session's directory is in a git repository, a connected device's diff pane shows the diff of your uncommitted changes. The device requests the
* **Diff of your changes**: when the session's directory is in a git repository, a connected device's diff pane shows your changes. The device requests the diff over the connection
+1 −1 lines
Upstream edited 22 Sep 2026 17:05 UTC Recorded here 22 Sep 2026 17:07 UTC
claude-code
hooks
±changed
+3−2
claude-codechanged
hooks
* `PermissionRequest`
`PermissionRequest` supports `command`, `http`, `mcp_tool`, and `prompt` hooks but not `agent` hooks. If you configure an agent hook on this event, Claude Code skips it and the per
Agent-based hooks (`type: "agent"`) are like prompt-based hooks but with multi-turn tool access. Instead of a single LLM call, an agent hook spawns a subagent that can read files,
Agent-based hooks (`type: "agent"`) are like prompt-based hooks but with multi-turn tool access. Instead of a single LLM call, an agent hook spawns a subagent that can read files,
+3 −2 lines
Upstream edited 22 Sep 2026 17:04 UTC Recorded here 22 Sep 2026 17:07 UTC
claude-docs
third-party/claude-desktop/configuration-changelog
±changed
+4−0
claude-docschanged
third-party/claude-desktop/configuration-changelog
<Update label="v2.2553.13" description="2026-09-21">
No configuration changes in this release.
</Update>
+4 −0 lines
Upstream edited 22 Sep 2026 16:57 UTC Recorded here 22 Sep 2026 17:07 UTC
claude-docs
cowork/changelog
±changed
+22−0
claude-docschanged
cowork/changelog
<Update label="v2.2553.13" description="2026-09-21">
Bundled Claude Code version: 2.1.280.
**General**
+22 −0 lines
Upstream edited 22 Sep 2026 16:57 UTC Recorded here 22 Sep 2026 17:07 UTC
claude-code
memory
±changed
+10−9
claude-codechanged
memory
* You or your organization set [`disableAllHooks`](/docs/en/settings-reference#disableallhooks) or [`allowManagedHooksOnly`](/docs/en/settings-reference#allowmanagedhooksonly), or
* You disabled the built-in `agents-md` plugin in `/plugin`
| | `CLAUDE.md`
| :------------------------------------------------------------------------------------------------------------------------------------ | :-----------------------------------------
| `/memory` and the **Memory files** list in `/context` | Listed
+10 −9 lines
Upstream edited 22 Sep 2026 16:52 UTC Recorded here 22 Sep 2026 17:07 UTC
claude-code
sub-agents
±changed
+3−1
claude-codechanged
sub-agents
The second filter applies to subagents running in the background. Apart from `Agent` and `ExitPlanMode`, which follow the first filter's conditions wherever the subagent runs, a ba
The second filter applies to subagents running in the background. Apart from `Agent` and `ExitPlanMode`, which follow the first filter's conditions wherever the subagent runs, a ba
Before v2.1.280, background subagents couldn't use `LSP`.
+3 −1 lines
Upstream edited 22 Sep 2026 16:49 UTC Recorded here 22 Sep 2026 17:07 UTC
claude-code
errors
±changed
+1−1
claude-codechanged
errors
* For a tool that [background subagents drop](/docs/en/sub-agents#available-tools), such as `LSP`, remove the entry. To keep the tool, [turn fork mode off](/docs/en/sub-agents#turn
* For a tool that [background subagents drop](/docs/en/sub-agents#available-tools), such as `CronCreate`, remove the entry. To keep the tool, [turn fork mode off](/docs/en/sub-agen
+1 −1 lines
Upstream edited 22 Sep 2026 16:49 UTC Recorded here 22 Sep 2026 17:07 UTC
claude-code
changelog
±changed
+117−0
claude-codechanged
changelog
<Update label="2.1.280" description="September 22, 2026">
* Added Claude Opus 5.5 (`claude-opus-5-5`), now the default Opus model — 1M context, $4/$20 per Mtok with \$0.20/Mtok cache reads
* Added mouse support to more lists in fullscreen mode: the wheel scrolls the `/skills` list, and a skill's state options in `/plugin` can be clicked
* Added `CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH` to change the 2,048-character cap on MCP tool descriptions and server instructions for every MCP server in the session
+117 −0 lines
Upstream edited 22 Sep 2026 16:48 UTC Recorded here 22 Sep 2026 17:07 UTC
claude-code
fast-mode
±changed
+8−8
claude-codechanged
fast-mode
| Model | Input (MTok) | Output (MTok) |
| -------- | ------------ | ------------- |
| Opus 5.5 | \$8 | \$40 |
| Opus 5 | \$10 | \$50 |
| Opus 4.8 | \$10 | \$50 |
+8 −8 lines
Upstream edited 22 Sep 2026 16:44 UTC Recorded here 28 Sep 2026 23:37 UTC
claude-code
keybindings
±changed
+25−3
claude-codechanged
keybindings
| `confirm:yes` | Y, Enter | Confirm action
| `confirm:no` | N, Escape | Decline action
| `confirm:yes` | Enter | Confirm action
| `confirm:no` | Escape | Decline action
+25 −3 lines
Upstream edited 22 Sep 2026 16:44 UTC Recorded here 22 Sep 2026 16:47 UTC
claude-code
fast-mode
±changed
+1−1
claude-codechanged
fast-mode
* Type `/fast` and press Tab to toggle on or off
* Run `/fast`, press Space to toggle on or off, then press Enter to confirm
+1 −1 lines
Upstream edited 22 Sep 2026 16:44 UTC Recorded here 22 Sep 2026 16:47 UTC
api
test-and-evaluate/strengthen-guardrails/handle-streaming-refusals
±changed
+8−8
apichanged
test-and-evaluate/strengthen-guardrails/handle-streaming-refusals
"model": "claude-opus-5",
"model": "claude-opus-5-5",
model="claude-opus-5",
model="claude-opus-5-5",
model: "claude-opus-5",
+8 −8 lines
Recorded here 22 Sep 2026 16:37 UTC
api
test-and-evaluate/develop-tests
±changed
+72−72
apichanged
test-and-evaluate/develop-tests
model="claude-opus-5",
model="claude-opus-5-5",
model: "claude-opus-5",
model: "claude-opus-5-5",
Model = Model.ClaudeOpus5,
+72 −72 lines
Recorded here 22 Sep 2026 16:37 UTC
api
resources/overview
±changed
+4−0
apichanged
resources/overview
<Card title="Claude Opus 5.5 System Card" icon="file" href="https://www.anthropic.com/claude-opus-5-5-system-card">
Detailed documentation of Claude Opus 5.5.
</Card>
+4 −0 lines
Recorded here 22 Sep 2026 16:37 UTC
api
release-notes/system-prompts/overview
±changed
+2−0
apichanged
release-notes/system-prompts/overview
<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" />
+2 −0 lines
Recorded here 22 Sep 2026 16:37 UTC
api
release-notes/overview
±changed
+9−0
apichanged
release-notes/overview
### September 22, 2026
* We've launched **Claude Opus 5.5** (`claude-opus-5-5`), a model for long-running agentic coding and knowledge work. It has a [1M token context window](https://platform.claude.com
* On Claude Opus 5.5, thinking can't be disabled: `thinking: {"type": "disabled"}` and `thinking: {"type": "enabled", ...}` return a 400 error. Omit the `thinking` field and contro
+9 −0 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/sonnet-5/overview
±changed
+3−3
apichanged
models/sonnet-5/overview
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Moderate | Adaptive | `high` | May 2
| [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) | 1M | 128K | $4 / $20 | Moderate | Adaptive (always on) | `medium` | Jun 2
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4
* **Price / MTok:** Input / output, base price per million tokens. Batch API requests are 50% off; prompt caching reads cost 10% of the base input price (2.5% on Claude Fable 5.1 a
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5.5, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, an
+3 −3 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/sonnet-5/migration-guide
±changed
+9−9
apichanged
models/sonnet-5/migration-guide
"content": "Are there an infinite number of prime numbers such that n mod 4 == 3?"
"content": "Find all pairs of positive integers (x, y) such that x^2 - y^2 = 2024."
content: Are there an infinite number of prime numbers such that n mod 4 == 3?
content: Find all pairs of positive integers (x, y) such that x^2 - y^2 = 2024.
"content": "Are there an infinite number of prime numbers such that n mod 4 == 3?",
+9 −9 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/sonnet-4-6/overview
±changed
+3−3
apichanged
models/sonnet-4-6/overview
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Adaptive | `high` | May 20
| [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) | 1M | 128K | $4 / $20 | Adaptive (always on) | `medium` | Jun 20
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4
* **Price / MTok:** Input / output, base price per million tokens. Batch API requests are 50% off; prompt caching reads cost 10% of the base input price (2.5% on Claude Fable 5.1 a
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5.5, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, an
+3 −3 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/sonnet-4-5/overview
±changed
+3−3
apichanged
models/sonnet-4-5/overview
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Adaptive | `high` | May 2026
| [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) | 1M | 128K | $4 / $20 | Adaptive (always on) | `medium` | Jun 2026
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4
* **Price / MTok:** Input / output, base price per million tokens. Batch API requests are 50% off; prompt caching reads cost 10% of the base input price (2.5% on Claude Fable 5.1 a
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5.5, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, an
+3 −3 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/overview
±changed
+24−24
apichanged
models/overview
If 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://platf
If 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://p
| Feature | Claude Fable 5.1
| :-------------------------------------------------------------------------------------------------------- | :---------------------------------------------------------------------
| Description | For demanding reasoning and long-horizon agentic work
+24 −24 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/opus-5/whats-new-opus-5
±changed
+5−1
apichanged
models/opus-5/whats-new-opus-5
<Note>
Claude Opus 5.5 is the current Opus model. See [What's new in Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/whats-new-opus-5-5).
</Note>
On Claude Opus 5, `thinking: {"type": "disabled"}` is accepted only when the effort level is `high` or below. Setting `thinking: {"type": "disabled"}` with effort `xhigh` or `max`
+5 −1 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/opus-5/overview
±changed
+23−14
apichanged
models/opus-5/overview
description: "Claude Opus 5 at a glance: what it's for, model IDs on every platform, context window, output limits, pricing, availability, and the guides and resources for building
description: "Claude Opus 5 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 5.5 is the curren
**Latest.** Released July 24, 2026.
**Legacy.** Released July 24, 2026.
+23 −14 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/opus-4-8/overview
±changed
+8−8
apichanged
models/opus-4-8/overview
description: "Claude Opus 4.8 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 4.8 is a legacy
description: "Claude Opus 4.8 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 4.8 is a legacy
Although Claude Opus 4.8 is still available, you should consider migrating to Claude Opus 5 for improved performance. [See Claude Opus 5](https://platform.claude.com/docs/en/models
Although Claude Opus 4.8 is still available, you should consider migrating to Claude Opus 5.5 for improved performance. [See Claude Opus 5.5](https://platform.claude.com/docs/en/mo
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Adaptive | `high` | May 2026
+8 −8 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/opus-4-7/overview
±changed
+8−8
apichanged
models/opus-4-7/overview
description: "Claude Opus 4.7 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 4.7 is a legacy
description: "Claude Opus 4.7 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 4.7 is a legacy
Although Claude Opus 4.7 is still available, you should consider migrating to Claude Opus 5 for improved performance. [See Claude Opus 5](https://platform.claude.com/docs/en/models
Although Claude Opus 4.7 is still available, you should consider migrating to Claude Opus 5.5 for improved performance. [See Claude Opus 5.5](https://platform.claude.com/docs/en/mo
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Adaptive | `high` | May 2026
+8 −8 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/opus-4-6/overview
±changed
+8−8
apichanged
models/opus-4-6/overview
description: "Claude Opus 4.6 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 4.6 is a legacy
description: "Claude Opus 4.6 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 4.6 is a legacy
Although Claude Opus 4.6 is still available, you should consider migrating to Claude Opus 5 for improved performance. [See Claude Opus 5](https://platform.claude.com/docs/en/models
Although Claude Opus 4.6 is still available, you should consider migrating to Claude Opus 5.5 for improved performance. [See Claude Opus 5.5](https://platform.claude.com/docs/en/mo
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Adaptive | `high` | May 20
+8 −8 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/opus-4-5/overview
±changed
+8−8
apichanged
models/opus-4-5/overview
description: "Claude Opus 4.5 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 4.5 is a legacy
description: "Claude Opus 4.5 reference: lifecycle status, model IDs on every platform, context window, output limits, pricing, and migration resources. Claude Opus 4.5 is a legacy
Although Claude Opus 4.5 is still available, you should consider migrating to Claude Opus 5 for improved performance. [See Claude Opus 5](https://platform.claude.com/docs/en/models
Although Claude Opus 4.5 is still available, you should consider migrating to Claude Opus 5.5 for improved performance. [See Claude Opus 5.5](https://platform.claude.com/docs/en/mo
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Adaptive | `high` | May 2026
+8 −8 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/mythos-5/overview
±changed
+3−3
apichanged
models/mythos-5/overview
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Adaptive | `high` | May 2026
| [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) | 1M | 128K | $4 / $20 | Adaptive (always on) | `medium` | Jun 2026
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4
* **Price / MTok:** Input / output, base price per million tokens. Batch API requests are 50% off; prompt caching reads cost 10% of the base input price (2.5% on Claude Fable 5.1 a
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5.5, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, an
+3 −3 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/mythos-5-1/overview
±changed
+3−3
apichanged
models/mythos-5-1/overview
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Moderate | Adaptive | `high` | May 2
| [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) | 1M | 128K | $4 / $20 | Moderate | Adaptive (always on) | `medium` | Jun 2
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4
* **Price / MTok:** Input / output, base price per million tokens. Batch API requests are 50% off; prompt caching reads cost 10% of the base input price (2.5% on Claude Fable 5.1 a
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5.5, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, an
+3 −3 lines
Recorded here 22 Sep 2026 16:37 UTC
api
models/haiku-4-5/overview
±changed
+3−3
apichanged
models/haiku-4-5/overview
| [Claude Opus 5](https://platform.claude.com/docs/en/models/opus-5/overview) | 1M | 128K | $5 / $25 | Moderate | Adaptive | `high` | May 2
| [Claude Opus 5.5](https://platform.claude.com/docs/en/models/opus-5-5/overview) | 1M | 128K | $4 / $20 | Moderate | Adaptive (always on) | `medium` | Jun 2
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4
* **Price / MTok:** Input / output, base price per million tokens. Batch API requests are 50% off; prompt caching reads cost 10% of the base input price (2.5% on Claude Fable 5.1 a
* **Max output:** Synchronous Messages API limit. On the Message Batches API, Claude Opus 5.5, Claude Opus 5, Claude Sonnet 5, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, an
+3 −3 lines
Recorded here 22 Sep 2026 16:37 UTC
Nothing on this page matches that. Press Enter to search every page of the ledger.