36| **`opus`** | Uses the latest Opus model for complex reasoning tasks |36| **`opus`** | Uses the latest Opus model for complex reasoning tasks |
37| **`haiku`** | Uses the fast and efficient Haiku model for simple tasks |37| **`haiku`** | Uses the fast and efficient Haiku model for simple tasks |
38| **`sonnet[1m]`** | Uses Sonnet with a [1 million token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows#context-window-sizes-by-model) for long sessions. No effect when `sonnet` already resolves to Sonnet 5.5 or Sonnet 5 with their native 1M window |38| **`sonnet[1m]`** | Uses Sonnet with a [1 million token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows#context-window-sizes-by-model) for long sessions. No effect when `sonnet` already resolves to Sonnet 5.5 or Sonnet 5 with their native 1M window |
39| **`opus[1m]`** | Uses Opus with a [1 million token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows#context-window-sizes-by-model) for long sessions |39| **`opus[1m]`** | Uses Opus with a [1 million token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows#context-window-sizes-by-model) for long sessions. No effect when `opus` already resolves to Opus 4.7 or later with its native 1M window |
40| **`opusplan`** | Special mode that uses `opus` during plan mode, then switches to `sonnet` for execution |40| **`opusplan`** | Special mode that uses `opus` during plan mode, then switches to `sonnet` for execution |
41 41
42The `opus`, `sonnet`, and `haiku` aliases resolve to the newest version on the Anthropic API and to an earlier version on some other providers:42The `opus`, `sonnet`, and `haiku` aliases resolve to the newest version on the Anthropic API and to an earlier version on some other providers:
292* [Claude Tag](https://claude.com/docs/claude-tag/overview) sessions run in cloud environments but don't receive server-managed settings; in a [self-hosted environment](/docs/en/self-hosted-environments), they still read the managed settings file in the runner image. To set the model for those sessions, see [Choose the model for a scope](https://claude.com/docs/claude-tag/admins/customize#choose-the-model-for-a-scope) in the Claude Tag admin guide.292* [Claude Tag](https://claude.com/docs/claude-tag/overview) sessions run in cloud environments but don't receive server-managed settings; in a [self-hosted environment](/docs/en/self-hosted-environments), they still read the managed settings file in the runner image. To set the model for those sessions, see [Choose the model for a scope](https://claude.com/docs/claude-tag/admins/customize#choose-the-model-for-a-scope) in the Claude Tag admin guide.
293* Cowork, the agentic-work tab in the Claude Desktop app, runs its sessions on Claude Code but, by design, does not receive server-managed settings from the claude.ai admin console. When the `availableModels` list in your server-managed settings is non-empty and a user picks a model outside it, the server rejects that model for a remote Cowork session. A managed settings file applies to Cowork sessions when it is present where the session runs; remote Cowork sessions run on Anthropic-managed VMs, where a device-deployed file is not present.293* Cowork, the agentic-work tab in the Claude Desktop app, runs its sessions on Claude Code but, by design, does not receive server-managed settings from the claude.ai admin console. When the `availableModels` list in your server-managed settings is non-empty and a user picks a model outside it, the server rejects that model for a remote Cowork session. A managed settings file applies to Cowork sessions when it is present where the session runs; remote Cowork sessions run on Anthropic-managed VMs, where a device-deployed file is not present.
294* Sessions on [third-party providers](/docs/en/server-managed-settings#platform-availability) such as Amazon Bedrock, Google Cloud's Agent Platform, Microsoft Foundry, and [Claude Platform on AWS](/docs/en/claude-platform-on-aws) do not receive server-managed settings, so deliver the allowlist through MDM or managed settings files there.294* Sessions on [third-party providers](/docs/en/server-managed-settings#platform-availability) such as Amazon Bedrock, Google Cloud's Agent Platform, Microsoft Foundry, and [Claude Platform on AWS](/docs/en/claude-platform-on-aws) do not receive server-managed settings, so deliver the allowlist through MDM or managed settings files there.
295* Server-managed delivery also requires the session to authenticate with an [eligible login or key](/docs/en/server-managed-settings#platform-availability). Fleets that generate keys only through an [`apiKeyHelper`](/docs/en/settings-reference#apikeyhelper) script should deliver the allowlist through MDM or managed settings files.295* Delivery from the admin console also requires the session to fetch with an [eligible login](/docs/en/server-managed-settings#platform-availability) to your organization or an OAuth token issued for it. For fleets that authenticate with API keys, whether configured directly or generated by an [`apiKeyHelper`](/docs/en/settings-reference#apikeyhelper) script, deliver the allowlist through MDM or managed settings files.
296* The Desktop Code tab also hosts [SSH sessions](/docs/en/desktop#ssh-sessions), which read the managed settings file from the remote host they run on. See [Desktop managed settings](/docs/en/desktop#managed-settings).296* The Desktop Code tab also hosts [SSH sessions](/docs/en/desktop#ssh-sessions), which read the managed settings file from the remote host they run on. See [Desktop managed settings](/docs/en/desktop#managed-settings).
297* The model pickers on claude.ai and in the Desktop app hide or grey out models excluded by your organization's allowlist. The picker state is a convenience for users; it doesn't enforce the allowlist.297* The model pickers on claude.ai and in the Desktop app hide or grey out models excluded by your organization's allowlist. The picker state is a convenience for users; it doesn't enforce the allowlist.
298 298
516 516
517This section covers content-based fallback from Fable models, Opus 5.5, Sonnet 5.5, and Opus 5. For availability-based fallback when a model is overloaded or unavailable, see [Fallback model chains](#fallback-model-chains).517This section covers content-based fallback from Fable models, Opus 5.5, Sonnet 5.5, and Opus 5. For availability-based fallback when a model is overloaded or unavailable, see [Fallback model chains](#fallback-model-chains).
518 518
519Fable models, Opus 5.5, Sonnet 5.5, and Opus 5 run with safety classifiers, which most often flag cybersecurity and biology content. When a classifier flags a request and the flagged category has a fallback model, Claude Code re-runs the request on that model and shows a notice in the transcript. For those two categories, the fallback model depends on which model refused:519Fable models, Opus 5.5, Sonnet 5.5, and Opus 5 run with safety classifiers, which most often flag cybersecurity and biology content. For those two categories, the fallback model depends on which model refused:
520 520
521* **Fable 5.1, Fable 5, and Opus 5.5**: biology-flagged requests re-run on Opus 5, and cybersecurity-flagged requests re-run on Opus 4.8.521* **Fable 5.1, Fable 5, and Opus 5.5**: biology-flagged requests re-run on Opus 5, and cybersecurity-flagged requests re-run on Opus 4.8.
522* **Sonnet 5.5**: cybersecurity-flagged requests re-run on Sonnet 5. Biology-flagged requests end with a refusal instead, because Sonnet 5.5 has no biology fallback model.522* **Sonnet 5.5**: cybersecurity-flagged requests re-run on Sonnet 5. Biology-flagged requests end with a refusal instead, because Sonnet 5.5 has no biology fallback model.
524 524
525On Amazon Bedrock, Google Cloud's Agent Platform, and Microsoft Foundry, Claude Code resolves these targets through your deployment's model IDs instead. See [Enable fallback on Bedrock, Agent Platform, and Foundry](#enable-fallback-on-bedrock-agent-platform-and-foundry).525On Amazon Bedrock, Google Cloud's Agent Platform, and Microsoft Foundry, Claude Code resolves these targets through your deployment's model IDs instead. See [Enable fallback on Bedrock, Agent Platform, and Foundry](#enable-fallback-on-bedrock-agent-platform-and-foundry).
526 526
527When Claude Code switches a flagged request to the fallback model for its category, it re-runs the request on that model. In your main conversation, it shows a notice in the transcript. To be asked first, see [Ask before switching](#ask-before-switching).
528
527After a fallback, the session continues on the fallback model. To return to your original model, run [`/model`](#setting-your-model).529After a fallback, the session continues on the fallback model. To return to your original model, run [`/model`](#setting-your-model).
528 530
529Category-based fallback requires Claude Code v2.1.219 or later. Before v2.1.219, every flagged Fable 5 request re-ran on your provider's default Opus model, and Opus 5 was not a fallback source.531Category-based fallback requires Claude Code v2.1.219 or later. Before v2.1.219, every flagged Fable 5 request re-ran on your provider's default Opus model, and Opus 5 was not a fallback source.
532 534
533#### Effort level after a fallback535#### Effort level after a fallback
534 536
535When Claude Code switches your session to the fallback model, it keeps the effort level the flagged request ran at in place of that model's default effort. For example, a session on Opus 5.5 at its default `medium` that falls back to Opus 4.8 stays at `medium`, although Opus 4.8 defaults to `high`.537When Claude Code switches your session to the fallback model, it keeps the effort level the flagged request ran at. For example, a session on Opus 5.5 at its default `medium` that falls back to Opus 4.8 stays at `medium`, although Opus 4.8 defaults to `high`.
536 538
537A different level applies in cases such as these:539A different level applies in cases such as these:
538 540
539* **Settings or organization default**: a level in your settings that applies to the fallback model, or a default effort your organization set for it, applies instead.
540* **Your own change**: once you choose an effort level, pick a model in `/model`, or resume the session later, the flagged request's level no longer carries over.541* **Your own change**: once you choose an effort level, pick a model in `/model`, or resume the session later, the flagged request's level no longer carries over.
541* **Skill effort**: a level that a skill's `effort` frontmatter set for the flagged request applies to that turn, and later turns run at the level the [effort resolution order](#adjust-effort-level) gives the fallback model.542* **Skill effort**: a level that a skill's `effort` frontmatter set for the flagged request applies to that turn, and later turns run at the level the [effort resolution order](#adjust-effort-level) gives the fallback model.
542 543
543The session header shows the level in effect next to the model name. To change it, run `/effort` in the session.544In the session, run `/effort status` to see the level in effect, or `/effort` to change it.
544 545
545#### Check what triggered fallback546#### Check what triggered fallback
546 547
550 551
551#### Ask before switching552#### Ask before switching
552 553
553To decide what happens each time a request is flagged, rather than switching automatically, run `/config` and turn off **Switch models when a message is flagged**, or set [`switchModelsOnFlag`](/docs/en/settings-reference#switchmodelsonflag) to `false` in your settings file. A flagged request then pauses the session with two options: switch to the fallback model, or edit the prompt and retry on the current model.554To decide what happens each time a request is flagged, run `/config`, select **Switch models when a message is flagged**, and choose **Ask each time**. You can also set [`switchModelsOnFlag`](/docs/en/settings-reference#switchmodelsonflag) to `false` in your settings file. Claude Code then pauses at a flagged request that would switch models and gives you two options: switch to the fallback model, or edit the prompt and retry.
555
556The first time a flagged request would switch models in an interactive session, Claude Code may ask whether to switch automatically from then on. It asks only if you haven't set `switchModelsOnFlag`, and it saves your choice as that key in your user settings.
557
558If you choose to stay on the current model instead, the saved value is `false`, the same as **Ask each time**. If you dismiss the question, Claude Code saves nothing and asks again the next time a flagged request would switch models.
554 559
555Some cases behave differently:560When you've chosen **Ask each time**, some cases behave differently:
556 561
557* When the flagged category has no fallback model, such as a biology flag on Opus 5 or Sonnet 5.5, Claude Code doesn't show the prompt and the request ends with the refusal.562* When the flagged category has no fallback model, such as a biology flag on Opus 5 or Sonnet 5.5, Claude Code doesn't show the prompt and the request ends with the refusal.
558* If both models flag the same request, you can edit the prompt and retry, or start a new session.563* If both models flag the same request, you can edit the prompt and retry, or start a new session.
559* In [cloud sessions](/docs/en/claude-code-on-the-web) on the mobile app, editing and retrying is not supported. Switch models, or continue the session from a desktop browser or the desktop app.564* In [cloud sessions](/docs/en/claude-code-on-the-web) on the mobile app, editing and retrying is not supported. Switch models, or continue the session from a desktop browser or the desktop app.
560* In [non-interactive mode](/docs/en/cli-reference#cli-flags) and SDK integrations that can't show the prompt, a flagged request ends the turn with a refusal instead.565* In [non-interactive mode](/docs/en/cli-reference#cli-flags) and SDK integrations that can't show the prompt, a flagged request ends the turn with a refusal instead.
566* In a [subagent](/docs/en/sub-agents), Claude Code doesn't show the prompt, and a flagged request that would switch models re-runs on the fallback model.
561* When the fallback target is blocked by [`availableModels`](#restrict-model-selection), Claude Code doesn't show the prompt. The flagged request ends with the refusal, the same as automatic fallback when the target is blocked.567* When the fallback target is blocked by [`availableModels`](#restrict-model-selection), Claude Code doesn't show the prompt. The flagged request ends with the refusal, the same as automatic fallback when the target is blocked.
562 568
563#### Enable fallback on Bedrock, Agent Platform, and Foundry569#### Enable fallback on Bedrock, Agent Platform, and Foundry
574* **Every source model**: set `ANTHROPIC_DEFAULT_OPUS_MODEL` to an Opus model ID to turn fallback on and give the flagged categories a target. A pin that names a model outside the Opus family, or the model that refused, leaves the refusal standing.580* **Every source model**: set `ANTHROPIC_DEFAULT_OPUS_MODEL` to an Opus model ID to turn fallback on and give the flagged categories a target. A pin that names a model outside the Opus family, or the model that refused, leaves the refusal standing.
575* **Sonnet 5.5**: in addition to the Opus pin, set `ANTHROPIC_DEFAULT_SONNET_MODEL` or keep a Sonnet 5 entry in the provider's model list to supply the model the request re-runs on. A Sonnet pin that names a model outside the Sonnet family, or Sonnet 5.5 itself, leaves the refusal standing.581* **Sonnet 5.5**: in addition to the Opus pin, set `ANTHROPIC_DEFAULT_SONNET_MODEL` or keep a Sonnet 5 entry in the provider's model list to supply the model the request re-runs on. A Sonnet pin that names a model outside the Sonnet family, or Sonnet 5.5 itself, leaves the refusal standing.
576 582
583The fallback model must also have a context window at least as large as the session's, or Claude Code doesn't switch and the flagged request ends with the same refusal. On these providers the source models run with the [1M context window](#extended-context) by default. Pin a model that also does, such as Opus 4.8 in `ANTHROPIC_DEFAULT_OPUS_MODEL` or Sonnet 5 in `ANTHROPIC_DEFAULT_SONNET_MODEL`, by an ID Claude Code [can match to that model](#pin-models-for-third-party-deployments).
584
577#### Security research and biology workloads585#### Security research and biology workloads
578 586
579Workloads in offensive security or biology, including penetration testing, Capture the Flag (CTF) exercises, and biology-adjacent codebases, trigger fallback frequently, often on the first request. For substantive biology work on Fable 5.1, Fable 5, or Opus 5.5, Claude Code moves the session to Opus 5 at the first flagged request, and later biology-flagged requests end in refusals there, because Opus 5 has no biology fallback. On Opus 5 and Sonnet 5.5, you get those refusals from the first flagged request.587Workloads in offensive security or biology, including penetration testing, Capture the Flag (CTF) exercises, and biology-adjacent codebases, trigger fallback frequently, often on the first request. For substantive biology work on Fable 5.1, Fable 5, or Opus 5.5, the first flagged request that switches models moves the session to Opus 5, and later biology-flagged requests end in refusals there, because Opus 5 has no biology fallback. On Opus 5 and Sonnet 5.5, you get those refusals from the first flagged request.
580 588
581This is expected routing for these domains, not an account flag. If your organization needs Fable-class capability for this work, ask your Anthropic account team about trusted access programs.589This is expected routing for these domains, not an account flag. If your organization needs Fable-class capability for this work, ask your Anthropic account team about trusted access programs.
582 590
598 606
5991. An explicit choice: the [`CLAUDE_CODE_EFFORT_LEVEL`](/docs/en/env-vars#variables) environment variable, launching with `--effort`, or `/effort` in the session ([a non-interactive `/effort` has narrower effect](#non-interactive-effort))6071. An explicit choice: the [`CLAUDE_CODE_EFFORT_LEVEL`](/docs/en/env-vars#variables) environment variable, launching with `--effort`, or `/effort` in the session ([a non-interactive `/effort` has narrower effect](#non-interactive-effort))
6002. Your settings: the level you saved for the model or an [`effortLevel`](/docs/en/settings-reference#effortlevel) key, with the precedence between them and across settings files stated at [`modelSettings`](/docs/en/settings-reference#modelsettings)6082. Your settings: the level you saved for the model or an [`effortLevel`](/docs/en/settings-reference#effortlevel) key, with the precedence between them and across settings files stated at [`modelSettings`](/docs/en/settings-reference#modelsettings)
6013. The model's default effort: `high` on every model that supports effort, except that Opus 5.5, Sonnet 5.5, and Haiku 5.5 default to `medium`, Opus 4.7 defaults to `xhigh`, and, when your organization sets a default effort level for its [organization default model](#organization-default-model), that level is the default when you run that model. After an automatic model fallback, see [Effort level after a fallback](#effort-level-after-a-fallback) for the level that applies.6093. The model's default effort: `high` on every model that supports effort, except that Opus 5.5, Sonnet 5.5, and Haiku 5.5 default to `medium`, Opus 4.7 defaults to `xhigh`, and, when your organization sets a default effort level for its [organization default model](#organization-default-model), that level is the default when you run that model
610
611After an automatic model fallback, see [Effort level after a fallback](#effort-level-after-a-fallback) for the level that applies.
602 612
603Opus 5.5 starts at `medium` unless one of the sources above sets a level for it, and a top-level `effortLevel` in your user settings file doesn't count for Opus 5.5. That key is the older form `/effort` wrote before Claude Code saved levels per model: it keeps applying where it applied before, on Opus 5, Fable 5.1, and earlier models, while Opus 5.5 and models released after it start at their own default until you choose a level for them with `/effort` or the `/model` picker. A top-level `effortLevel` in project, local, or managed settings, or one passed with `--settings`, applies to every model.613Opus 5.5 starts at `medium` unless one of the sources above sets a level for it, and a top-level `effortLevel` in your user settings file doesn't count for Opus 5.5. That key is the older form `/effort` wrote before Claude Code saved levels per model: it keeps applying where it applied before, on Opus 5, Fable 5.1, and earlier models, while Opus 5.5 and models released after it start at their own default until you choose a level for them with `/effort` or the `/model` picker. A top-level `effortLevel` in project, local, or managed settings, or one passed with `--settings`, applies to every model.
604 614
709 719
710<a id="extended-context-with-1m" />720<a id="extended-context-with-1m" />
711 721
722<span id="sonnet-5-5-and-sonnet-5-context-window" />
723
712### Extended context724### Extended context
713 725
714Fable 5.1, Fable 5, Sonnet 5 and later, Haiku 5.5, Opus 4.6 and later, and Sonnet 4.6 support a [1 million token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows#context-window-sizes-by-model) for long sessions with large codebases.726Fable 5.1, Fable 5, Sonnet 5 and later, Haiku 5.5, Opus 4.6 and later, and Sonnet 4.6 support a [1 million token context window](https://platform.claude.com/docs/en/build-with-claude/context-windows#context-window-sizes-by-model) for long sessions with large codebases.
715 727
716On the Anthropic API, Fable 5.1, Fable 5, Sonnet 5 and later, Haiku 5.5, and Opus 4.7 and later run with the 1M window on every plan, including Pro. You don't select a `[1m]` variant or turn on usage credits for the 1M window on these models. Fable usage itself can bill to usage credits on some plans; see [Fable and usage credits](#fable-and-usage-credits).728Fable 5.1, Fable 5, Sonnet 5 and later, Haiku 5.5, and Opus 4.7 and later run with the 1M window by default, with no `[1m]` suffix needed. That includes sessions on Amazon Bedrock, Google Cloud's Agent Platform, and Microsoft Foundry, and [Claude apps gateway](/docs/en/claude-apps-gateway) sessions. To run them with a 200K window instead, see [Turn off 1M context](#turn-off-1m-context).
717 729
718Opus 4.6 and Sonnet 4.6 reach 1M only through their `[1m]` variant, and access to that variant depends on your plan. On Max, Team, and Enterprise plans, including both Team Standard and Team Premium seats, Opus 4.6 with 1M context is included with your subscription. Sonnet 4.6 with 1M context requires [usage credits](https://support.claude.com/en/articles/12429409-extra-usage-for-paid-claude-plans) on every subscription plan, including Max.730Opus 4.6 and Sonnet 4.6 reach 1M only through their `[1m]` variant, and access to that variant depends on your plan. On Max, Team, and Enterprise plans, including both Team Standard and Team Premium seats, Opus 4.6 with 1M context is included with your subscription. Sonnet 4.6 with 1M context requires [usage credits](https://support.claude.com/en/articles/12429409-extra-usage-for-paid-claude-plans) on every subscription plan, including Max.
719 731
725 737
726Claude Code checks these plan requirements only when it connects to the Anthropic API directly. If you point `ANTHROPIC_BASE_URL` at an [LLM gateway](/docs/en/llm-gateway#subscriptions-and-gateways) and your saved claude.ai login stays the active credential, Claude Code doesn't check your plan's usage credits. The `[1m]` options stay available in `/model`, and the gateway decides whether the request succeeds. Before v2.1.229, Claude Code rejected `/model sonnet[1m]` in that configuration when it couldn't confirm usage credits on the account.738Claude Code checks these plan requirements only when it connects to the Anthropic API directly. If you point `ANTHROPIC_BASE_URL` at an [LLM gateway](/docs/en/llm-gateway#subscriptions-and-gateways) and your saved claude.ai login stays the active credential, Claude Code doesn't check your plan's usage credits. The `[1m]` options stay available in `/model`, and the gateway decides whether the request succeeds. Before v2.1.229, Claude Code rejected `/model sonnet[1m]` in that configuration when it couldn't confirm usage credits on the account.
727 739
728<span id="context-window-behind-a-gateway" />740On the Anthropic API, the 1M context window uses standard model pricing with no premium for tokens beyond 200K, except on Haiku 5.5, which [costs more on prompts longer than 100K tokens](#haiku-5-5-context-window-and-pricing). For plans where extended context is included with your subscription, usage remains covered by your subscription. For plans that access extended context through usage credits, tokens are billed to usage credits.
729
730If you set `ANTHROPIC_BASE_URL` to an [LLM gateway](/docs/en/llm-gateway) or another proxy, Claude Code gives each model it recognizes the same context window the model has on the Anthropic API. Fable 5.1, Fable 5, Sonnet 5 and later, Haiku 5.5, and Opus 4.7 and later get the 1M window with no `[1m]` variant to select, and a model that reaches 1M only through its `[1m]` variant, such as Opus 4.6, runs at 200K without it. Claude Code can't detect a lower limit that the gateway or the server behind it enforces. If your gateway rejects requests above 200K tokens, set [`CLAUDE_CODE_AUTO_COMPACT_WINDOW=200000`](/docs/en/env-vars) in the environment that starts Claude Code, so sessions on every model [compact at that boundary](#set-the-auto-compact-window).
731 741
732To turn off 1M context, set `CLAUDE_CODE_DISABLE_1M_CONTEXT=1`. Claude Code removes 1M model variants from the model picker. On models with a native 1M window, such as Sonnet 5 and the Fable models, it also treats the model as having a 200K context window:742#### Select 1M context for Opus 4.6 or Sonnet 4.6
733
734* With auto-compaction on, sessions compact at the 200K boundary through [auto-compaction](#set-the-auto-compact-window). Setting the auto-compact window above 200K doesn't lift the hold, because Claude Code caps that window at the model's context window.
735* With auto-compaction off, sessions stop at the 200K boundary with the [context-limit error](/docs/en/errors#prompt-is-too-long) instead of compacting.
736 743
737Before v2.1.223, Claude Code held only Sonnet 5, Opus 4.8, and Opus 5 sessions to 200K. See [environment variables](/docs/en/env-vars).744To select a 1M variant by name, append the `[1m]` suffix to a model alias or a full model name:
738
739The 1M context window uses standard model pricing with no premium for tokens beyond 200K, except on Haiku 5.5, which [costs more on prompts longer than 100K tokens](#haiku-5-5-context-window-and-pricing). For plans where extended context is included with your subscription, usage remains covered by your subscription. For plans that access extended context through usage credits, tokens are billed to usage credits.
740
741If your account supports 1M context, the option appears in the `/model` picker in the latest versions of Claude Code. If you don't see it, restart your session, and on a third-party provider check whether your deployment [pinned the model](#pin-models-for-third-party-deployments) with an `ANTHROPIC_DEFAULT_*_MODEL` variable.
742
743You can also use the `[1m]` suffix with model aliases or full model names:
744 745
745```text theme={null}746```text theme={null}
746# Use the opus[1m] or sonnet[1m] alias747# Append [1m] to a full model name
747/model opus[1m]748/model claude-opus-4-6[1m]
748/model sonnet[1m]749/model claude-sonnet-4-6[1m]
749 750
750# Or append [1m] to a full model name751# Or to an alias: the suffix applies to the model the alias resolves to
751/model claude-opus-4-8[1m]752/model opus[1m]
752```753```
753 754
754#### Sonnet 5.5 and Sonnet 5 context window755<span id="context-window-behind-a-gateway" />
755 756
756On the Anthropic API, Sonnet 5.5 and Sonnet 5 always run with the 1M context window. There is no 200K variant, no `[1m]` suffix to select, and no usage credits required on any plan. Sessions auto-compact before the window fills, at about 967K tokens by default; set [`CLAUDE_CODE_AUTO_COMPACT_WINDOW`](/docs/en/env-vars) to choose a different threshold.757#### Context window behind an LLM gateway
757 758
758Claude Code gives Sonnet 5.5 and Sonnet 5 the same 1M window behind an [LLM gateway](/docs/en/llm-gateway) or another custom `ANTHROPIC_BASE_URL`. If your gateway enforces a lower limit, see [the context window behind a gateway](#context-window-behind-a-gateway).759If you set `ANTHROPIC_BASE_URL` to an [LLM gateway](/docs/en/llm-gateway) or another proxy, Claude Code gives each model it recognizes the same context window the model has on the Anthropic API. Fable 5.1, Fable 5, Sonnet 5 and later, Haiku 5.5, and Opus 4.7 and later get the 1M window with no `[1m]` variant to select, and a model that reaches 1M only through its `[1m]` variant, such as Opus 4.6, runs at 200K without it. Claude Code can't detect a lower limit that the gateway or the server behind it enforces. If your gateway rejects requests above 200K tokens, set [`CLAUDE_CODE_AUTO_COMPACT_WINDOW=200000`](/docs/en/env-vars) in the environment that starts Claude Code, so sessions on every model [compact at that boundary](#set-the-auto-compact-window).
759 760
760This setting budgets the window at 200K instead:761#### Turn off 1M context
761 762
762* **`CLAUDE_CODE_DISABLE_1M_CONTEXT=1`**: holds sessions on every model with a native 1M window to a 200K window; see [Extended context](#extended-context) for how the hold is enforced. Useful for deployments that need to cap context.763To keep sessions at a 200K window, set `CLAUDE_CODE_DISABLE_1M_CONTEXT=1` in your shell or in a [settings file](/docs/en/env-vars#set-environment-variables). Claude Code removes the `[1m]` model variants from the model picker. On models that run with the 1M window by default, such as the Fable models, Sonnet 5 and later, and Opus 4.7 and later, it also treats the model as having a 200K context window:
764
765* With auto-compaction on, sessions compact at the 200K boundary through [auto-compaction](#set-the-auto-compact-window). Setting the auto-compact window above 200K doesn't lift the hold, because Claude Code caps that window at the model's context window.
766* With auto-compaction off, sessions stop at the 200K boundary with the [context-limit error](/docs/en/errors#prompt-is-too-long) instead of compacting.
763 767
764#### Haiku 5.5 context window and pricing768#### Haiku 5.5 context window and pricing
765 769
795If you don't set an auto-compact window, Claude Code compacts when the conversation reaches the model's context limit, except in these sessions:799If you don't set an auto-compact window, Claude Code compacts when the conversation reaches the model's context limit, except in these sessions:
796 800
797* [Cloud sessions](/docs/en/claude-code-on-the-web) compact as the conversation approaches the model's limit801* [Cloud sessions](/docs/en/claude-code-on-the-web) compact as the conversation approaches the model's limit
798* Sonnet 4.6 and Opus 4.6 without [extended context](#extended-context) compact at the 200K boundary, and so do Opus 4.8 and later when they run with a 200K context window, such as on Amazon Bedrock, Google Cloud's Agent Platform, and Microsoft Foundry802* Sonnet 4.6 and Opus 4.6 without [extended context](#extended-context) compact at the 200K boundary
799* When you set [`CLAUDE_CODE_DISABLE_1M_CONTEXT=1`](/docs/en/env-vars), models with a native 1M window, such as Sonnet 5 and the Fable models, compact at the 200K boundary803* When you set [`CLAUDE_CODE_DISABLE_1M_CONTEXT=1`](/docs/en/env-vars), models with a native 1M window, such as Sonnet 5 and the Fable models, compact at the 200K boundary
800* Models running with a native 1M window compact before the window fills, at about 967K tokens by default. On the Anthropic API, these include Sonnet 5, Haiku 5.5, the Fable models, and Opus 4.7 and later. On Amazon Bedrock, Google Cloud's Agent Platform, and Microsoft Foundry, see [Pin models for third-party deployments](#pin-models-for-third-party-deployments) for which models run with that window. Behind a custom `ANTHROPIC_BASE_URL`, see [the context window behind a gateway](#context-window-behind-a-gateway)804* Models running with a native 1M window compact before the window fills, at about 967K tokens by default. These include the Fable models, Sonnet 5 and later, Haiku 5.5, and Opus 4.7 and later. Behind a custom `ANTHROPIC_BASE_URL`, see [the context window behind a gateway](#context-window-behind-a-gateway)
801* Sessions on a model ID Claude Code doesn't recognize, such as an [LLM gateway](/docs/en/llm-gateway) alias, compact at the context window Claude Code assumes for the ID; see [Correct the window for a gateway or custom model ID](#correct-the-window-for-a-gateway-or-custom-model-id)805* Sessions on a model ID Claude Code doesn't recognize, such as an [LLM gateway](/docs/en/llm-gateway) alias, compact at the context window Claude Code assumes for the ID; see [Correct the window for a gateway or custom model ID](#correct-the-window-for-a-gateway-or-custom-model-id)
802 806
803### Correct the window for a gateway or custom model ID807### Correct the window for a gateway or custom model ID
891 895
892Apply the same pattern for `ANTHROPIC_DEFAULT_FABLE_MODEL`, `ANTHROPIC_DEFAULT_SONNET_MODEL`, and `ANTHROPIC_DEFAULT_HAIKU_MODEL`. For current and legacy model IDs across all providers, see [Models overview](https://platform.claude.com/docs/en/about-claude/models/overview). To upgrade users to a new model version, update these environment variables and redeploy.896Apply the same pattern for `ANTHROPIC_DEFAULT_FABLE_MODEL`, `ANTHROPIC_DEFAULT_SONNET_MODEL`, and `ANTHROPIC_DEFAULT_HAIKU_MODEL`. For current and legacy model IDs across all providers, see [Models overview](https://platform.claude.com/docs/en/about-claude/models/overview). To upgrade users to a new model version, update these environment variables and redeploy.
893 897
894To enable [extended context](#extended-context) for a pinned model, append `[1m]` to the model ID in `ANTHROPIC_DEFAULT_OPUS_MODEL`, `ANTHROPIC_DEFAULT_SONNET_MODEL`, or `ANTHROPIC_DEFAULT_FABLE_MODEL`:898A pinned model with a native 1M window, such as Opus 4.8 or Sonnet 5, runs with the [1M context window](#extended-context) without any suffix when Claude Code can match the pinned ID to that model. The ID matches when it contains the model's Anthropic API ID, as `us.anthropic.claude-opus-4-8` contains `claude-opus-4-8`, or when a [`modelOverrides`](#override-model-ids-per-version) entry maps the model to it. On a pinned ID that Claude Code can't match to a model, sessions run with a 200K window by default unless the ID carries the `[1m]` suffix.
899
900For a model that reaches 1M through its `[1m]` variant, such as Opus 4.6 or Sonnet 4.6, enable extended context by appending `[1m]` to the model ID in `ANTHROPIC_DEFAULT_OPUS_MODEL` or `ANTHROPIC_DEFAULT_SONNET_MODEL`:
895 901
896```bash theme={null}902```bash theme={null}
897export ANTHROPIC_DEFAULT_OPUS_MODEL='claude-opus-4-8[1m]'903export ANTHROPIC_DEFAULT_OPUS_MODEL='claude-opus-4-6[1m]'
898```904```
899 905
900With the `[1m]` suffix, the 1M context window applies to all usage of the pinned alias, including the plan-mode Opus phase of [`opusplan`](#opusplan-model-setting) and [subagents](/docs/en/sub-agents#choose-a-model) whose `model` frontmatter names the alias.906With the `[1m]` suffix, the 1M context window applies to all usage of the pinned alias, including the plan-mode Opus phase of [`opusplan`](#opusplan-model-setting) and [subagents](/docs/en/sub-agents#choose-a-model) whose `model` frontmatter names the alias.
901 907
902* Claude Code strips the suffix before sending the model ID to your provider.908* Claude Code strips the suffix before sending the model ID to your provider.
903* Only append `[1m]` when the underlying model [supports 1M context](https://platform.claude.com/docs/en/build-with-claude/context-windows#context-window-sizes-by-model).909* Only append `[1m]` when the underlying model [supports 1M context](https://platform.claude.com/docs/en/build-with-claude/context-windows#context-window-sizes-by-model).
904* The suffix is read per variable, not per model. On Amazon Bedrock, Google Cloud's Agent Platform, and Microsoft Foundry, a model ID without `[1m]` in one variable uses 200K context even if another variable sets the same model with the suffix. Sonnet 5 always runs with the 1M window on these providers and never needs the suffix.910* The suffix is read per variable, not per model. On Amazon Bedrock, Google Cloud's Agent Platform, and Microsoft Foundry, an Opus 4.6 or Sonnet 4.6 ID without `[1m]` in one variable uses 200K context even if another variable sets the same model with the suffix.
905 911
906When you set an `ANTHROPIC_DEFAULT_*_MODEL` variable, the `/model` picker shows one row for that model in place of the family's built-in rows, including any 1M context rows. To reach the 1M window without adding the suffix to that variable, your users run `/model opus[1m]`, and Claude Code applies the suffix to the model the variable names. `/model sonnet[1m]` works the same way.912When you set an `ANTHROPIC_DEFAULT_*_MODEL` variable, the `/model` picker shows one row for that model in place of the family's built-in rows, including any 1M context rows. To reach the 1M window without adding the suffix to that variable, your users run `/model opus[1m]`, and Claude Code applies the suffix to the model the variable names. `/model sonnet[1m]` works the same way.
907 913
