Claude model for demanding reasoning and long-horizon agentic workClaude Fable 5.2 by anthropic is offered by 32 providers on this page. Its official list price is $8.00 per 1M input tokens and $40.00 per 1M output tokens. The lowest paid channel is AIGROW at $3.20 / $16.00 per 1M, about 2.5× below the list price. That channel is a third-party gateway, so confirm its availability and rate limits before depending on it.
Artificial Analysis rates it 64.6 on the Intelligence Index, with 72.6 for Engineering, its strongest domain. Against its lowest blended price of $6.40 per 1M, that is roughly 10 index points per dollar, which is the value ratio the leaderboards rank on. Median output speed is 87 tokens per second, with 206.99s to the first token. Running one task of the Artificial Analysis suite costs about $26.5386, which reflects how many tokens its reasoning consumes rather than the unit price alone.
The context window is 1,000,000 tokens, with an output limit of 128,000 tokens. It supports reasoning, tool use, and structured output. Accepted input modalities are Text, Image, and PDF. Its training knowledge cuts off at 2026-06.
Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.
| Provider | Tier | Input | Output | Cache read | Cache write | Context | Output limit | Status |
|---|---|---|---|---|---|---|---|---|
| AIGROW claude-fable-5.2 | Gateway | $3.20 | $16.00 | $0.20 | $6.40 | 1,000,000 | 128,000 | |
| Outer Labs | Gateway | $4.80 | $24.00 | $0.12 | $6.00 | 1,000,000 | 128,000 | |
| AnthropicOfficial | First-party | $8.00 | $40.00 | $0.20 | $10.00 | 1,000,000 | 128,000 | |
| Amazon Bedrock global.anthropic.claude-fable-5-2 | Cloud | $8.00 | $40.00 | $0.20 | $10.00 | 1,000,000 | 128,000 | |
| Amazon Bedrock anthropic.claude-fable-5-2 | Cloud | $8.00 | $40.00 | $0.20 | $10.00 | 1,000,000 | 128,000 | |
| Merge Gateway | Gateway | $8.00 | $40.00 | $0.20 | $10.00 | 1,000,000 | 128,000 | |
| Kilo Gateway anthropic/claude-fable-5.2 | Gateway | $8.00 | $40.00 | $0.20 | $10.00 | 1,000,000 | 128,000 | |
| DevPass (LLM Gateway) | Gateway | $8.00 | $40.00 | $0.20 | $10.00 | 1,000,000 | 128,000 | |
| Vertex claude-fable-5-2@default | Cloud | $8.00 | $40.00 | $0.20 | $10.00 | 1,000,000 | 128,000 | |
| DigitalOcean anthropic-claude-fable-5.2 | Cloud | $8.00 | $40.00 | $0.20 | $10.00 | 1,000,000 | 128,000 |
Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.
2 / 32 providers expose no reasoning control (reasoning_options: []).
| Mode | Provider | Input | Output | Cache read | Cache write |
|---|---|---|---|---|---|
| fast | Anthropic | $16.00 | $80.00 | $0.40 | $20.00 |
| fast | Vertex | $16.00 | $80.00 | $0.40 | $20.00 |
| fast | Vertex (Anthropic) | $16.00 | $80.00 | $0.40 | $20.00 |
| Name | Conditions | Score | Metric | Source |
|---|---|---|---|---|
| FrontierCode | variant: xhigh effort; safeguards on · v2.0 | 49.1 | score | Source ↗ |
| FrontierSWE | variant: highest available effort; safeguards on · v2 | 69.9 | score | Source ↗ |
| AutomationBench | variant: highest available effort; safeguards on | 51.4 | score | Source ↗ |
| Terminal-Bench-Science | variant: highest available effort; safeguards on · v0.1 | 74.6 | score | Source ↗ |
| WANDR | variant: highest available effort; safeguards on | 82.1 | relaxed score | Source ↗ |
| OSWorld | variant: highest available effort; safeguards on · dataset: offline set · v2.1 | 89.3 | partial score | Source ↗ |
| RoboDojo Sim | variant: max effort; internal run | 64.9 | average success rate | Source ↗ |
| AA-Briefcase | variant: highest available effort; safeguards on · v1.1 | 66.1 | rubric score | Source ↗ |
| AA-Briefcase | variant: highest available effort; safeguards on · v1.1 | 1785 | Elo | Source ↗ |
| Remote Labor Index | variant: high effort; safeguards on | 41.3 | score | Source ↗ |
| Humanity's Last Exam | variant: with tools; highest available effort; safeguards on | 74.2 | score | Source ↗ |
| Chartography | variant: no tools; highest available effort; safeguards on | 73.4 | score | Source ↗ |
The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.
Individual evaluations run by Artificial Analysis, on the same reasoning tier as the intelligence score above. Each benchmark has its own task set and scale (percent correct, points or an index), so rows are not comparable with one another.