Claude Fable
Anthropic’s peak widely released Claude (Fable 5.1) for long-horizon agents and deep reasoning.
Anthropic’s Claude Fable 5.1 — the most capable widely released Claude for long-horizon agents, deep reasoning, and demanding coding workflows. Mythos 5.1 is the same underlying model with relaxed cyber/life-sciences safeguards for trusted-access programs.
Why Claude Fable matters
Fable 5.1 is what teams reach for when Sonnet and Opus are not enough—long-running agents, hard research, and frontier coding. It is the public capability ceiling of the Claude family (Mythos 5.1 is trusted-access). Cache-read pricing is lower than Fable 5; list input/output rates are unchanged.
Vision · Tool calling · Thinking · MCP · Coding
Last reviewed: 27 September 2026
When to choose Claude Fable
Decision guidance for architects—not a feature list.
Best for
- Long-horizon agentic workflows
- Deep multi-step reasoning
- Hardest agentic coding tasks
- Peak-capability Claude deployments
Avoid if
- Sonnet or Opus already meets quality targets
- Cost and latency dominate over peak capability
Strengths
DataAIHub qualitative scores from 1 to 5, for architects. Not official benchmark results or a public ranking.
- Peak Claude capability★★★★★
- Long-horizon agents★★★★★
- Deep reasoning★★★★★
- Cost efficiency★★☆☆☆
- Open weights★☆☆☆☆
Ecosystem
Built by
- Anthropic
Fable is Anthropic’s peak widely released Claude capability tier.
Competes with
- GPT-5
Primary rival for peak closed-model agent and coding capability.
Works with
Recommended for
- AI Agents
Built for demanding long-horizon agentic workflows.
Often paired with
- Claude Opus
Teams often keep Opus for high-stakes work and escalate to Fable only when needed.
- Claude Sonnet
Sonnet handles everyday production traffic; Fable handles the hardest agent threads.
How Claude Fable evolved
Key moments in chronological order.
- API
Fable 5.1 effort changes keep the cache
Changing effort mid-session on Fable 5.1 does not reset the prompt cache. API keys and cloud providers can set a one-hour cache lifetime; forked subagents reuse the parent cache.
- Model
Successor to Fable 5 for long-running agents, coding, and research. API id claude-fable-5-1; 1M context, 128k max output, always-on adaptive thinking. High effort default in Claude Code; Medium in Cowork and Claude.ai.
- Product
Same underlying model as Fable 5.1 with more permissive cyber and life-sciences safeguards for Cyber Verification and Life Sciences Verification programs.
- Platform
Fable 5.1 cache reads $0.25/MTok
Cache reads cut 75% vs Fable 5. List price remains $10 input / $50 output per 1M tokens. Anthropic estimates ~25% lower typical-workload cost vs Fable 5, up to ~45% on cache-heavy agentic work.
- API
Computer use GA + browser use on Fable 5
Fable 5 supports GA computer_toolset_20260801 and browser_toolset_20260801 on the Claude API.
- Model
Claude Fable 5 launch
Most capable widely released Claude for long-horizon agents and deep reasoning.
- Product
Mythos 5 (Glasswing)
Limited-access peer of Fable for approved Project Glasswing customers.
- API
claude-fable-5 API
Generally available on Claude API and major cloud partners at premium rates.
- Product
Adaptive thinking + effort
Effort parameter replaces manual thinking budgets for depth control.
- Product
1M context default
Million-token context and large max outputs for long agent threads.
Overview
Capabilities
- Vision: Yes
- Audio input: No
- Tool calling: Yes
- Thinking: Yes
- MCP: Yes
- Coding: Yes
- Structured output: Yes
Technical specifications
- Provider
- Anthropic
- License
- Proprietary
- Context window
- 1M
- Parameters
- Undisclosed
- Architecture
- Transformer (proprietary)
- Release
- 2026-09
- Modalities
- Text, Image
- Vision
- Yes
- Audio input
- No
- Tool calling
- Yes
- Thinking
- Yes
- MCP
- Yes
- Open weights
- No
- API
- Yes
- Pricing (input)
- ~$10 per 1M tokens (claude-fable-5-1)
- Pricing (output)
- ~$50 per 1M tokens
Supported modalities
Text · Image
Context window
1M (1,000,000 tokens)
Pricing
List price matches Fable 5. Cache reads $0.25 / 1M tokens (75% below Fable 5). Anthropic estimates ~25% lower typical-workload cost vs Fable 5, up to ~45% on cache-heavy agentic work. Verify live pricing.
Availability
Claude API model ID: claude-fable-5-1 (1M context, 128k max output, always-on adaptive thinking). Also on AWS Bedrock, Google Cloud, and Microsoft Azure/Foundry. Defaults: High effort in Claude Code; Medium in Claude apps. As of 2026-09-16, chat and Cowork merge into one conversation on Pro/Max (Docs/Slides/Design beta from any chat). Messages API on-demand compaction is a 2026-09-14 beta (compact-2026-09-04). As of 2026-09-24, changing effort mid-session does not reset the prompt cache; API keys and cloud providers can set a one-hour cache lifetime, and forked subagents reuse the parent cache. Mythos 5.1 is trusted-access only (Cyber and Life Sciences Verification Programs). Covered Model with extra retention/safety policy. Enterprise Frontier Safeguards (customer-controlled review) announced for a phased fall rollout; eligible customers may use ZDR until EFS is available. SynthID-Text watermarking on models launched on/after 2026-08-02; detection API in private preview. Messages API version 2026-10-21, the default in new Anthropic SDK majors and ant CLI 2.0.0, returns 400 if a request sends temperature, top_p, or top_k. Keep anthropic-version 2023-06-01 when a call still needs those parameters. On Amazon Bedrock, 2026-10-21 is only for Claude 4.7 and later. Those new SDK and CLI releases are Apache-2.0; earlier releases stay MIT.
Use cases
- Long-horizon agentic workflows
- Deep multi-step reasoning and research
- Hard agentic coding and systems work
- Peak-capability Claude deployments
- Computer/browser automation at frontier capability
Strengths
- Peak widely released Claude capability
- Strong long-horizon agent performance
- GA computer use + browser toolsets
- 1M-token context with adaptive thinking
Limitations
- Highest Claude cost among widely released tiers
- Closed weights
- Overkill when Sonnet or Opus is enough
Related guides
Related benchmarks
Related research
- Claude Fable 5.1 and Mythos 5.1
Original Paper
Related GitHub
Related tools
Related rankings
Related comparisons
Companies
Explore more models
- Claude OpusAnthropicAnthropic’s Claude Opus 5.5 (API id claude-opus-5-5, released 2026-09-22) for complex agentic coding, enterprise work, long-context analysis, and careful instruction following. Anthropic says it matches Fable 5.1 on most work at a lower price. Fable 5.1 remains a separate higher list-price SKU. Sonnet 5.5 (claude-sonnet-5-5, 2026-09-28) is the current Sonnet at the same $2/$10 list price as Sonnet 5. Haiku 5.5 (claude-haiku-5-5, 2026-10-07) is the current Haiku.
- Claude SonnetAnthropicAnthropic’s Claude Sonnet 5.5 (API id claude-sonnet-5-5, released 2026-09-28) for well-scoped everyday coding, agents, and documents. List price matches Sonnet 5 at $2/$10 per 1M tokens. Opus 5.5 remains the high-stakes tier; Haiku 5.5 (2026-10-07) is the low-cost subagent and high-volume tier.
- Claude HaikuAnthropicAnthropic’s fast, cost-efficient Claude tier for high-volume chat, classification, extraction, and sub-agent steps where latency and price matter more than peak reasoning. Claude Haiku 5.5 (API id claude-haiku-5-5, released 2026-10-07) is the current Haiku: Anthropic’s fastest model at standard speed, the first Haiku with an adjustable effort setting, and around 75% cheaper to run than Haiku 4.5 on average. Sonnet 5.5 and Opus 5.5 remain better for complex agentic coding.
- Gemini 3.1 ProGoogleGoogle’s Pro-class Gemini for hard reasoning and native multimodal work. Prefer API id gemini-3.1-pro-preview; Gemini 3.5 Pro remains partner-testing. This slug is not Gemini 4 Argon (gemini-4-argon, public 2026-10-13). On the Gemini API, gemini-2.5-pro has no announced shutdown date and is not deprecated; access is limited to projects that have already used it.
- Gemini 4 ArgonGoogleStable Gemini API id gemini-4-argon, released 13 October 2026. Google’s model page describes it as the flagship for complex software engineering, deep mathematical and scientific problems, autonomous multi-step agents, and long decode. Thinking can be set to low, medium, or high. The long-decode guide names medium as the default for most coding and agentic work. It documents function calling, structured outputs, code execution, search grounding, URL context, and file search. Computer use is in preview. The model page does not document native MCP support. Image generation, audio generation, and the Live API are not supported.
- GPT-5.6OpenAIOpenAI’s GPT-5.6 family (Sol flagship, Terra balanced, Luna cost-efficient) for complex reasoning, coding, multimodal understanding, and agentic tool use. The gpt-5.6 API alias routes to Sol. GPT-6.1 Sol (gpt-6.1-sol, 2026-09-29) is the current lower-cost Sol; GPT-6 Sol and Luna (2026-09-22) are earlier cost tiers. This entry remains GPT-5.6, which is still available.