By using Claude Fable 5, you consent to data retention for purposes of monitoring safe and compliant use of AI. Anthropic's data retention policy does not allow zero data retention. Learn moreOpens in new tab
Azure, Google Vertex, and Amazon Bedrock endpoints are BYOK-only while we work on enabling public access.
Claude Fable 5 is a Mythos-class model from Anthropic, built for autonomous knowledge work and coding. It supports text, image, and file inputs with text output, with reasoning support and a 1M-token context window. It is suited for long-running, complex, and asynchronous tasks that previously required frequent human check-ins.
It is particularly strong at end-to-end work that would otherwise take a person hours, days, or weeks - taking on problems that are long-running, ambiguous, or highly multi-step. It executes well-scoped tasks with few mistakes, automatically self-correcting through verification loops, and ships with robust safeguards.
Providers
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).
BYOK only | $5.00 | $25.00 | $0.50 | -- | -- |
|---|
- Input $5.00 / Output $25.00-- tok/s-- uptimeBYOK only
Pricing
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Weighted average
Claude Fable 5 (batch) — Price history
| $0.00 | $0.00 | 0.0% | 100.0% |
- Effective in $0.00 / Effective out $0.00Token share100.0%
Performance
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
Uptime
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Not enough uptime data to display yet.
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.
Benchmarks
Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.
| Source | Benchmark | Score |
|---|---|---|
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) AutomationBench-AA | 54.1% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) GDP.pdf | 24.0% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) AA-AnalystAgent | 48.7% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) EnterpriseOps-Gym-AA | 51.1% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) GPQA Diamond | 92.6% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) HLE | 55.5% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) IFBench | 63.5% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) τ²-Bench Telecom | 98.5% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) AA-LCR | 82.3% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) τ³-Banking | 38.1% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) GDPval-AA | 55.9% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) CritPt | 28.6% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) SciCode | 61.0% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) Terminal-Bench Hard | 62.9% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) Terminal-Bench 4.0 | 42.4% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) Terminal-Bench 2.1 | 84.6% |
| Artificial Analysis | Claude Fable 5 (Max, Opus 4.8 Fallback) MLCR-AA | 20.6% |
| Design Arena | Claude Fable 5 Agents Arena Agenticgamedev Elo | 1231 |
| Design Arena | Claude Fable 5 Agents Arena Agentichtmlslides Elo | 1254 |
| Design Arena | Claude Fable 5 Agents Arena Agenticslides(Html) Elo | 1252 |
| Design Arena | Claude Fable 5 Agents Arena Androidnative Elo | 1276 |
| Design Arena | Claude Fable 5 Agents Arena Full Stack Elo | 1253 |
| Design Arena | Claude Fable 5 Agents Arena Godotgamedev Elo | 1263 |
| Design Arena | Claude Fable 5 Agents Arena Htmlslides Elo | 1174 |
| Design Arena | Claude Fable 5 Agents Arena Mobile Apps Elo | 1254 |
| Design Arena | Claude Fable 5 Agents Arena Python-Pptxslides Elo | 1268 |
| Design Arena | Claude Fable 5 Agents Arena Webapps Elo | 1242 |
| Design Arena | Claude Fable 5 Models Arena 3D Elo | 1297 |
| Design Arena | Claude Fable 5 Models Arena Asciiart Elo | 1310 |
| Design Arena | Claude Fable 5 Models Arena Code Categories Elo | 1299 |
| Design Arena | Claude Fable 5 Models Arena Data Visualization Elo | 1303 |
| Design Arena | Claude Fable 5 Models Arena Game Development Elo | 1324 |
| Design Arena | Claude Fable 5 Models Arena SVG Elo | 1292 |
| Design Arena | Claude Fable 5 Models Arena UI Component Elo | 1307 |
| Design Arena | Claude Fable 5 Models Arena Website Elo | 1287 |
| Vals AI | Claude Fable 5 Code Migration | 55.1% |
| Vals AI | Claude Fable 5 CorpFin v2 | 71.8% |
| Vals AI | Claude Fable 5 EMB | 73.7% |
| Vals AI | Claude Fable 5 Finance Agent (v2) | 56.3% |
| Vals AI | Claude Fable 5 GPQA Diamond | 93.2% |
| Vals AI | Claude Fable 5 Harvey's Legal Agent Benchmark v1 | 11.3% |
| Vals AI | Claude Fable 5 IOI v1 | 72.3% |
| Vals AI | Claude Fable 5 LiveCodeBench | 89.8% |
| Vals AI | Claude Fable 5 LegalBench | 88.6% |
| Vals AI | Claude Fable 5 Legal Research Bench | 49.5% |
| Vals AI | Claude Fable 5 MedCode | 56.1% |
| Vals AI | Claude Fable 5 MedScribe | 88.5% |
| Vals AI | Claude Fable 5 MMLU Pro | 91.5% |
| Vals AI | Claude Fable 5 MMMU Pro | 89.3% |
| Vals AI | Claude Fable 5 MortgageTax | 68.9% |
| Vals AI | Claude Fable 5 ProgramBench | 2.0% |
| Vals AI | Claude Fable 5 ProofBench v1.1 | 95.0% |
| Vals AI | Claude Fable 5 Public Benefits Bench v1.1 | 70.4% |
| Vals AI | Claude Fable 5 Vals RSI Index | 24.3% |
| Vals AI | Claude Fable 5 SAGE v1 | 51.9% |
| Vals AI | Claude Fable 5 SWE-bench | 95.0% |
| Vals AI | Claude Fable 5 Tax Agent Bench | 38.3% |
| Vals AI | Claude Fable 5 TaxEval v2 | 76.9% |
| Vals AI | Claude Fable 5 Terminal-Bench 4.0 | 41.4% |
| Vals AI | Claude Fable 5 Terminal-Bench-Science | 15.7% |
| Vals AI | Claude Fable 5 Vals Index | 61.4% |
| Vals AI | Claude Fable 5 Vals Multimodal Index | 74.2% |
| Vals AI | Claude Fable 5 Vibe Code Bench v1.1 | 90.4% |
| Vals AI | Claude Fable 5 Vals Web Search Index | 46.9% |
Apps
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
Not enough data for apps using this model yet.
Activity
Token volume and request traffic to this model over time.
API
Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.
Anthropic: Claude Fable versions
Not currently available through EU in-region routing. In-region routing docs
Frequently asked questions
Claude Fable 5 is a Mythos-class model from Anthropic, built for autonomous knowledge work and coding. It supports text, image, and file inputs with text output, with reasoning support and a 1M-token context window. It is suited for long-running, complex, and asynchronous tasks that previously required frequent human check-ins.
Claude Fable 5 (batch) costs $5.00/M input tokens and $25.00/M output tokens, with separate rates for Cache Read at $0.50/M tokens, Cache Write at $6.25/M tokens, Cache Write (1h) at $10.00/M tokens, and Web Search at $10.00/1K calls.
Claude Fable 5 (batch) has a 1,000,000 token context window. It supports up to 128,000 completion tokens.
Yes. Claude Fable 5 (batch) accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
Claude Fable 5 (batch) accepts text, images, and files such as PDFs as input and returns text.
Claude Fable 5 (batch) is served by 1 provider on OpenRouter: Anthropic.
Claude Fable 5.2, Claude Haiku 5.5, Claude Sonnet 5.5 and 12 more are other text models from Anthropic.
Claude Fable 5 (batch) was released on June 9, 2026.