feat(models): add Claude Fable 5.1 and align Anthropic prompt-cache minimums (#7380)

Author: waleedlatif1

apps/docs/content/docs/workflows/blocks/agent.mdx

@@ -109,7 +109,7 @@ Live tool-call chips stream for **OpenAI, Anthropic, Azure Anthropic, Google, Ve
 | Provider | Streamed thinking | Models |
 |----------|-------------------|--------|
 | OpenAI | Summaries only — Requires OpenAI organization verification; falls back to no summaries. | `gpt-5.6-sol`, `gpt-5.6-terra`, `gpt-5.6-luna`, `gpt-5.5-pro`, `gpt-5.5`, `gpt-5.4-pro`, `gpt-5.4`, `gpt-5.4-mini`, `gpt-5.4-nano`, `gpt-5.2-pro`, `gpt-5.2`, `gpt-5.1`, `gpt-5-pro`, `gpt-5`, `gpt-5-mini`, `gpt-5-nano`, `o4-mini`, `o3`, `o3-mini`, `o1` |
-| Anthropic | Summaries only — These generations omit full thinking; Sim requests summarized thinking on streaming runs. | `claude-fable-5`, `claude-sonnet-5`, `claude-opus-5`, `claude-opus-4-8`, `claude-opus-4-7`, `claude-opus-4-6`, `claude-sonnet-4-6`, `claude-opus-4-5`, `claude-opus-4-1`, `claude-sonnet-4-5`, `claude-haiku-4-5` |
+| Anthropic | Summaries only — These generations omit full thinking; Sim requests summarized thinking on streaming runs. | `claude-fable-5-1`, `claude-fable-5`, `claude-sonnet-5`, `claude-opus-5`, `claude-opus-4-8`, `claude-opus-4-7`, `claude-opus-4-6`, `claude-sonnet-4-6`, `claude-opus-4-5`, `claude-opus-4-1`, `claude-sonnet-4-5`, `claude-haiku-4-5` |
 | Azure OpenAI | Summaries only — Requires OpenAI organization verification; falls back to no summaries. | `azure/gpt-5.4`, `azure/gpt-5.4-mini`, `azure/gpt-5.4-nano`, `azure/gpt-5.2`, `azure/gpt-5.1`, `azure/gpt-5.1-codex`, `azure/gpt-5`, `azure/gpt-5-mini`, `azure/gpt-5-nano`, `azure/o3`, `azure/o4-mini` |
 | Azure Anthropic | Summaries only — These generations omit full thinking; Sim requests summarized thinking on streaming runs. | `azure-anthropic/claude-opus-4-6`, `azure-anthropic/claude-opus-4-5`, `azure-anthropic/claude-sonnet-4-5`, `azure-anthropic/claude-opus-4-1`, `azure-anthropic/claude-haiku-4-5` |
 | Google | Summaries only | `gemini-3.6-flash`, `gemini-3.5-flash-lite`, `gemini-3.5-flash`, `gemini-3.1-pro-preview`, `gemini-3.1-flash-lite`, `gemini-3-flash-preview`, `gemini-2.5-pro`, `gemini-2.5-flash`, `gemini-2.5-flash-lite` |

apps/sim/providers/anthropic/core.request.test.ts

@@ -248,3 +248,58 @@ describe('executeAnthropicProviderRequest prompt caching', () => {
     expect(payload.system).toBeUndefined()
   })
 })
+
+describe('executeAnthropicProviderRequest forced tool use', () => {
+  const forcedTool = {
+    id: 'publish',
+    name: 'publish',
+    description: 'Publish a post',
+    params: {},
+    parameters: { type: 'object', properties: {}, required: [] },
+    usageControl: 'force' as const,
+  }
+
+  const textReply = {
+    id: 'msg-answer',
+    type: 'message',
+    role: 'assistant',
+    model: 'claude-fable-5-1',
+    content: [{ type: 'text', text: 'Done' }],
+    stop_reason: 'end_turn',
+    stop_sequence: null,
+    usage: { input_tokens: 2, output_tokens: 2 },
+  }
+
+  async function runWithForcedTool(model: string) {
+    const create = vi.fn().mockResolvedValue({ ...textReply, model })
+    const warn = vi.fn()
+    await executeAnthropicProviderRequest(
+      {
+        model,
+        apiKey: 'test-key',
+        maxTokens: 1024,
+        messages: [{ role: 'user', content: 'Publish this' }],
+        tools: [forcedTool],
+      },
+      {
+        providerId: 'anthropic',
+        providerLabel: 'Anthropic',
+        createClient: () => ({ messages: { create } }) as never,
+        logger: { info: vi.fn(), warn, error: vi.fn(), debug: vi.fn() },
+      }
+    )
+    return { payload: create.mock.calls[0][0] as Anthropic.Messages.MessageCreateParams, warn }
+  }
+
+  it('forces the tool on models that accept forced tool_choice', async () => {
+    const { payload } = await runWithForcedTool('claude-sonnet-4-5')
+    expect(payload.tool_choice).toEqual({ type: 'tool', name: 'publish' })
+  })
+
+  it('drops forced tool_choice on Claude Fable 5.1 because the API rejects it', async () => {
+    const { payload, warn } = await runWithForcedTool('claude-fable-5-1')
+    expect(payload.tools?.map((tool) => tool.name)).toEqual(['publish'])
+    expect(payload).not.toHaveProperty('tool_choice')
+    expect(warn).toHaveBeenCalledWith(expect.stringContaining('rejects forced tool_choice'))
+  })
+})

apps/sim/providers/anthropic/core.thinking.test.ts

@@ -13,6 +13,7 @@ import { describeModelLevel } from '@/providers/utils'
 describe('buildThinkingConfig', () => {
   it('requests summarized display for omitted-display models on agent-events runs', () => {
     for (const model of [
+      'claude-fable-5-1',
       'claude-fable-5',
       'claude-sonnet-5',
       'claude-opus-5',
@@ -27,6 +28,7 @@ describe('buildThinkingConfig', () => {
 
   it('never adds display on legacy runs (no agent events)', () => {
     for (const model of [
+      'claude-fable-5-1',
       'claude-fable-5',
       'claude-sonnet-5',
       'claude-opus-5',

apps/sim/providers/anthropic/core.ts

@@ -132,7 +132,7 @@ const ANTHROPIC_THINKING_OUTPUT_HEADROOM = 4096
 
 /**
  * Checks if a model supports adaptive thinking (thinking.type: "adaptive").
- * Fable 5 supports ONLY adaptive thinking (always on; type: "disabled" is rejected).
+ * Fable 5 and Fable 5.1 support ONLY adaptive thinking (always on; type: "disabled" is rejected).
  * Sonnet 5 supports ONLY adaptive thinking (manual budget_tokens returns a 400 error).
  * Opus 5, Opus 4.8, and Opus 4.7 support ONLY adaptive thinking (no extended thinking / budget_tokens).
  * Opus 4.6 and Sonnet 4.6 support both extended and adaptive thinking — use adaptive.
@@ -155,10 +155,21 @@ function supportsAdaptiveThinking(modelId: string): boolean {
   )
 }
 
+/**
+ * Claude Fable 5.1 and Claude Mythos 5.1 reject forced tool use: a `tool_choice` of
+ * type `tool` or `any` returns a 400 (`tool_choice: type "tool" and "any" are not
+ * supported for this model.`). Thinking is always on for these models, so a forced
+ * call would skip it. The request is sent with the default `auto` instead.
+ */
+function rejectsForcedToolChoice(modelId: string): boolean {
+  const normalizedModel = modelId.toLowerCase()
+  return normalizedModel.includes('fable-5-1') || normalizedModel.includes('mythos-5-1')
+}
+
 /**
  * Builds the thinking configuration for the Anthropic API based on model capabilities and level.
  *
- * - Fable 5, Sonnet 5, Opus 5, Opus 4.8, Opus 4.7: Uses adaptive thinking only (no extended thinking support)
+ * - Fable 5.1, Fable 5, Sonnet 5, Opus 5, Opus 4.8, Opus 4.7: Uses adaptive thinking only (no extended thinking support)
  * - Opus 4.6, Sonnet 4.6: Uses adaptive thinking with effort parameter
  * - Other models: Uses budget_tokens-based extended thinking
  *
@@ -417,7 +428,13 @@ export async function executeAnthropicProviderRequest(
     } else if (toolChoice === 'none') {
       payload.tool_choice = { type: 'none' }
     } else if (toolChoice !== 'auto') {
-      payload.tool_choice = toolChoice
+      if (rejectsForcedToolChoice(request.model)) {
+        logger.warn(
+          `Model ${modelId} rejects forced tool_choice; sending tool "${toolChoice.name}" with tool_choice auto`
+        )
+      } else {
+        payload.tool_choice = toolChoice
+      }
     }
   }
 

apps/sim/providers/models.test.ts

@@ -61,8 +61,8 @@ describe('prompt caching capability', () => {
 
   it('reports the vendor minimum prefix, raised for Haiku', () => {
     expect(getPromptCachingMinimumTokens('claude-sonnet-5')).toBe(1024)
-    expect(getPromptCachingMinimumTokens('claude-haiku-4-5')).toBe(2048)
-    expect(getPromptCachingMinimumTokens('azure-anthropic/claude-haiku-4-5')).toBe(2048)
+    expect(getPromptCachingMinimumTokens('claude-haiku-4-5')).toBe(4096)
+    expect(getPromptCachingMinimumTokens('azure-anthropic/claude-haiku-4-5')).toBe(4096)
     expect(getPromptCachingMinimumTokens('gpt-5.5')).toBeNull()
   })
 })

apps/sim/providers/models.ts

@@ -825,6 +825,27 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
       promptCaching: { minimumCacheableTokens: 1024 },
     },
     models: [
+      {
+        id: 'claude-fable-5-1',
+        pricing: {
+          input: 10.0,
+          cachedInput: 0.25,
+          output: 50.0,
+          updatedAt: '2026-09-01',
+        },
+        capabilities: {
+          nativeStructuredOutputs: true,
+          maxOutputTokens: 128000,
+          promptCaching: { minimumCacheableTokens: 512 },
+          thinking: {
+            levels: ['low', 'medium', 'high', 'xhigh', 'max'],
+            default: 'high',
+            streamed: 'summary',
+          },
+        },
+        contextWindow: 1000000,
+        releaseDate: '2026-09-01',
+      },
       {
         id: 'claude-fable-5',
         pricing: {
@@ -836,6 +857,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
         capabilities: {
           nativeStructuredOutputs: true,
           maxOutputTokens: 128000,
+          promptCaching: { minimumCacheableTokens: 512 },
           thinking: {
             levels: ['low', 'medium', 'high', 'xhigh', 'max'],
             default: 'high',
@@ -877,6 +899,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
         capabilities: {
           nativeStructuredOutputs: true,
           maxOutputTokens: 128000,
+          promptCaching: { minimumCacheableTokens: 512 },
           thinking: {
             levels: ['low', 'medium', 'high', 'xhigh', 'max'],
             default: 'high',
@@ -918,6 +941,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
         capabilities: {
           nativeStructuredOutputs: true,
           maxOutputTokens: 128000,
+          promptCaching: { minimumCacheableTokens: 2048 },
           thinking: {
             levels: ['low', 'medium', 'high', 'xhigh', 'max'],
             default: 'high',
@@ -939,6 +963,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
           temperature: { min: 0, max: 1 },
           nativeStructuredOutputs: true,
           maxOutputTokens: 128000,
+          promptCaching: { minimumCacheableTokens: 4096 },
           thinking: {
             levels: ['low', 'medium', 'high', 'max'],
             default: 'high',
@@ -981,6 +1006,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
           temperature: { min: 0, max: 1 },
           nativeStructuredOutputs: true,
           maxOutputTokens: 64000,
+          promptCaching: { minimumCacheableTokens: 4096 },
           thinking: {
             levels: ['low', 'medium', 'high'],
             default: 'high',
@@ -1086,7 +1112,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
           temperature: { min: 0, max: 1 },
           nativeStructuredOutputs: true,
           maxOutputTokens: 64000,
-          promptCaching: { minimumCacheableTokens: 2048 },
+          promptCaching: { minimumCacheableTokens: 4096 },
           thinking: {
             levels: ['low', 'medium', 'high'],
             default: 'high',
@@ -1457,6 +1483,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
           temperature: { min: 0, max: 1 },
           nativeStructuredOutputs: true,
           maxOutputTokens: 128000,
+          promptCaching: { minimumCacheableTokens: 4096 },
           thinking: {
             levels: ['low', 'medium', 'high', 'max'],
             default: 'high',
@@ -1478,6 +1505,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
           temperature: { min: 0, max: 1 },
           nativeStructuredOutputs: true,
           maxOutputTokens: 64000,
+          promptCaching: { minimumCacheableTokens: 4096 },
           thinking: {
             levels: ['low', 'medium', 'high'],
             default: 'high',
@@ -1541,7 +1569,7 @@ export const PROVIDER_DEFINITIONS: Record<string, ProviderDefinition> = {
           temperature: { min: 0, max: 1 },
           nativeStructuredOutputs: true,
           maxOutputTokens: 64000,
-          promptCaching: { minimumCacheableTokens: 2048 },
+          promptCaching: { minimumCacheableTokens: 4096 },
           thinking: {
             levels: ['low', 'medium', 'high'],
             default: 'high',