Data as of October 17, 2026 · How the score is built
Released Sep 1, 20261M contextClaude Fable 5.1 and Claude Mythos 5.1 system card
Claude Fable 5.1
Decision readingClaude Fable 5.1 scores 83.2 out of 100 and ranks #4 of 224. This profile shows 33 source-displayable benchmark rows; its strongest eligible category is Agentic at #2. API pricing is $10 input and $50 output per million tokens, with cached input at $0.25.
Released Sep 1, 2026 — see all recent releases
Decision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Capability
83.2/100
field median 49.6#4 of 224 ranked models
Public
#4of 224
Verified #4 of 88
Price
$10input / $50 output
input median $0.75cached $0.25 · blended $30
Speed
71tok/s
field median 92 tok/sFirst token 258.47 s
Context
1Mtokens
field median 256,000Maximum output length is tracked separately
Strongest published evidence
Agentic ranks #2. Particularly useful for coding agents, browser research, and computer-use workflows.
Validate before choosing
33 published rows leave some tracked benchmark slots empty. Reasoning is its lowest eligible category at #4.
Source-linked · 33 displayable benchmark rows
Category score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank #2 of 128Percentile 99thWeight 22%12 benchmarksVerified | 80.9 | 12 benchmarks | Verified | |||
| CodingRank #3 of 150Percentile 99thWeight 20%11 benchmarksVerified | 76.8 | 11 benchmarks | Verified | |||
| ReasoningRank #4 of 28Percentile 89thWeight 17%3 benchmarksVerified | 85.2 | 3 benchmarks | Verified | |||
| MultimodalWeight 12%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| KnowledgeRank #2 of 184Percentile 99thWeight 12%4 benchmarksVerified | 86.9 | 4 benchmarks | Verified | |||
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| Inst. FollowingWeight 5%1 benchmarkVerified | Score pending | 1 benchmark | Verified | |||
| MathRank Not rankedWeight 5%2 benchmarksVerified | 96.8 | 2 benchmarks | Verified |
33 of 681 tracked benchmark slots have displayable evidence · bars run 0–100
Coverage detailsHow much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic12/12 verified
- Coding11/11 verified
- Reasoning3/3 verified
- MultimodalNot measured
- Knowledge4/4 verified
- MultilingualNot measured
- Inst. Following1/1 verified
- Math2/2 verified
Capability shape
Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.
Claude Fable 5.1 category percentile values
- Agentic99th percentile
- Coding99th percentile
- Reasoning89th percentile
- MultimodalNot eligible
- Knowledge99th percentile
- MultilingualNot eligible
- Instruction followingNot eligible
- MathNot eligible
The dashed outline is median of 6 nearest peers.
Eligible category ranks
- Agentic#2/128
- Coding#3/150
- Reasoning#4/28
- MultimodalNot ranked
- Knowledge#2/184
- MultilingualNot ranked
- Inst. FollowingNot ranked
- MathNot ranked
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding11 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| SWE-bench Pro | Score81.2% | Versus best verified row Best verified: Claude Opus 5.5 · 89.9% | Gap8.7 behind | Weight26% ref. weight | |
| DeepSWE | Score67.4% | Versus best verified row Best verified: Muse Wonder Full · 79.0% | Gap11.6 behind | Weight15% ref. weight | |
| CursorBench 3.2 | Score73.4% | Versus best verified row Best verified: Claude Fable 5.1 · 73.4% | GapBest verified | Weight10% ref. weight | Benchmark exact |
| FrontierSWE v2 | Score56.3% | Versus best verified row Best verified: GPT-6 Astra · 65.5% | Gap9.2 behind | Weight8% ref. weight | Benchmark exact |
| LiveCodeBench (Vals)LiveCodeBench, Vals AI run | Score90.5% | Versus best verified row Best verified: Claude Fable 5.1 · 90.5% | GapBest verified | Weight8% ref. weight | |
| SWE Multilingual | Score89.1% | Versus best verified row Best verified: Claude Opus 5.5 · 93.9% | Gap4.8 behind | Weight5% ref. weight | |
| Bug Hunt Bench | Score43.0 fixes | Versus best verified row Best verified: Claude Sonnet 5.5 · 51.3 fixes | Gap8.3 behind | WeightDisplay only | Benchmark exact |
| SWE MultimodalSWE-bench Multimodal | Score54.7% | Versus best verified row Best verified: Claude Opus 5.5 · 61.4% | Gap6.7 behind | WeightDisplay only | |
| ProgramBenchProgramBench: Can Language Models Rebuild Programs From Scratch? | Score87.6% | Versus best verified row Best verified: Claude Opus 5 · 93.0% | Gap5.4 behind | WeightDisplay only | |
| CursorBench 4.0 | Score51.8% | Versus best verified row Best verified: Claude Opus 5.5 · 57.8% | Gap6 behind | WeightDisplay only | Benchmark exact |
| PostTrainBench v1.1 | Score40.2% | Versus best verified row Best verified: Claude Opus 5.5 · 49.3% | Gap9.1 behind | WeightDisplay only | Provider exact |
Agentic12 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Terminal-Bench 4.0 | Score55.80% | Versus best verified row Best verified: Claude Sonnet 5.5 · 70.60% | Gap14.8 behind | Weight8% ref. weight | |
| AutomationBench | Score31.4% | Versus best verified row Best verified: DeepSeek V4.1 Flash 1017 · 60.5% | Gap29.1 behind | Weight6% ref. weight | |
| Terminal-Bench 2.1 (Vals)Terminal-Bench 2.1, Vals AI run | Score85.0% | Versus best verified row Best verified: GPT-6 Astra · 87.3% | Gap2.3 behind | Weight4% ref. weight | |
| Toolathlon-Verified | Score77.8% | Versus best verified row Best verified: Claude Opus 5 · 80.6% | Gap2.8 behind | Weight3% ref. weight | |
| Terminal-Bench-Science 0.1 | Score52.6% | Versus best verified row Best verified: GPT-6 Astra · 64.6% | Gap12 behind | WeightDisplay only | |
| OSWorld 2.0 | Score41.7% | Versus best verified row Best verified: GPT-6 Astra · 72.6% | Gap30.9 behind | WeightDisplay only | |
| Toolathlon Verified Pass@3 | Score81.5% | Versus best verified row Best verified: Claude Opus 5 · 87.0% | Gap5.5 behind | WeightDisplay only | |
| Toolathlon Verified Pass³Toolathlon Verified Pass cubed | Score73.1% | Versus best verified row Best verified: Claude Opus 5 · 73.1% | GapBest verified | WeightDisplay only | |
| Toolathlon Verified avg. turnsToolathlon Verified average assistant turns | Score23.7 turns | Versus best verified row Best verified: Claude Sonnet 5.5 · 31.6 turns | Gap7.9 behind | WeightDisplay only | |
| ApprenticeBenchApprenticeBench: end-to-end computer use, continual learning, and long-horizon agency on a real accounts-payable job | Score72% | Versus best verified row Best verified: Claude Fable 5.1 · 72% | GapBest verified | WeightDisplay only | Benchmark exact |
| CWE-bench v1 | Score58.0% | Versus best verified row Best verified: GPT-6 Astra · 68.0% | Gap10 behind | WeightDisplay only | Benchmark exact |
| MazeBenchMazeBench: visual spatial reasoning in a 3D open world | Score10% | Versus best verified row Best verified: GPT-6 Astra · 23% | Gap13 behind | WeightDisplay only | Benchmark exact |
Reasoning3 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| ARC-AGI-2Abstraction and Reasoning Corpus for AGI v2 | Score90% | Versus best verified row Best verified: GPT-6 Astra · 95% | Gap5 behind | WeightWeighted 25% | |
| ARC-AGI-1ARC-AGI-1 Semi-Private Evaluation | Score97.50% | Versus best verified row Best verified: GPT-6 Astra · 98.50% | Gap1 behind | WeightDisplay only | |
| GraphWalks BFS 256K–1M | Score65.0% | Versus best verified row Best verified: Gemini 4 Argon · 84.2% | Gap19.2 behind | WeightDisplay only | Provider exact |
Knowledge4 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| HLEHumanity's Last Exam | Score65% | Versus best verified row Best verified: Claude Fable 5.1 · 65% | GapBest verified | Weight44% ref. weight | |
| HLE w/o toolsHumanity's Last Exam without tools | Score60.9% | Versus best verified row Best verified: Claude Opus 5.5 · 64.4% | Gap3.5 behind | Weight7% ref. weight | |
| MMLU-Pro (Vals)MMLU-Pro, Vals AI run | Score92.4% | Versus best verified row Best verified: Claude Fable 5.1 · 92.4% | GapBest verified | Weight6% ref. weight | Verified |
| GPQA Diamond (Vals)GPQA Diamond, Vals AI run | Score93.4% | Versus best verified row Best verified: Gemini 3.1 Pro · 95.5% | Gap2.1 behind | Weight2% ref. weight |
Inst. Following1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Gray Swan IPI (15 attempts)Gray Swan Indirect Prompt Injection, attack success at 15 attempts | Score1.0% | Versus best verified row Best verified: Gemini 4 Argon · 0.7% | Gap0.3 behind | WeightDisplay only | Provider exact |
Math2 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| FrontierMath v2 (Tiers 1-3)FrontierMath v2 Tiers 1-3 | Score90.175% | Versus best verified row Best verified: GPT-6 Astra · 93.684% | Gap3.5 behind | WeightWeighted 30% | Benchmark exact |
| FrontierMath v2 (Tier 4)FrontierMath v2 Tier 4 | Score87.805% | Versus best verified row Best verified: GPT-6.1 Sol · 100.000% | Gap12.2 behind | WeightWeighted 10% | Benchmark exact |
What it costs to get this score
Published API price against the public score. The x-axis uses a log scale; the dashed path marks models that are not beaten by a cheaper, higher-scoring option. Price uses average of published input and output rates.
Claude Fable 5.1 · 83.2 score · $30 blended per million tokens
The chart opens on the current model. Scroll horizontally to inspect the full price axis.
Horizontal: blended price per million tokens, log scale · Vertical: public score
Lineage
The sequence follows explicit supersedes links. Each score is estimated for that model; a relative can inform a sparse estimate but never sets a floor, so a newer release can score below an earlier one. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Base entry
Claude Fable 5 · 81.3Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- claude-fable-5-1Anthropic Fable 5.1 and Mythos 5.1 launch
- Context window
- 1MClaude Fable 5.1 and Claude Mythos 5.1 system card
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- Not sourced yetClaude Fable 5.1 and Claude Mythos 5.1 system card
- Output modalities
- Not sourced yetClaude Fable 5.1 and Claude Mythos 5.1 system card
- Parameters
- Not disclosed by the provider
- Availability
- Claude API · Claude products · AWS · Google Cloud · Microsoft AzureAnthropic Fable 5.1 and Mythos 5.1 launch
- Cloud regions
- Not tracked yet
- Lifecycle
- activeAnthropic Fable 5.1 and Mythos 5.1 launch
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Published at $0.25 per million cached input tokensAnthropic Fable 5.1 launch
- Self-host
- Weights are not published
- Rate limits
- Not tracked yet
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
Claude Fable 5.1 ranks #4 of 224 on the public leaderboard with a score of 83.18/100. Its source-verified position is #4 of 88.
Claude Fable 5.1 is a proprietary model with a 1M context window. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.
Generally available on the Claude API as `claude-fable-5-1` and across Anthropic, Amazon Web Services, Google Cloud, and Microsoft Azure. Fable 5.1 allows defensive vulnerability discovery, while exploit generation and some dual-use cyber work still route to Opus models. Eligible enterprise customers can use zero data retention until Enterprise Frontier Safeguards rolls out in phases beginning later in fall 2026.
Official exact-value snapshot from Anthropic's September 2026 Claude Fable 5.1 and Claude Mythos 5.1 launch page and system card, supplemented by Cursor's live CursorBench 3.2 table and Bug Hunt Bench's blind-graded production-repository evaluation. We store only exact protocol matches, including the highest published effort where several settings are shown. FrontierSWE v2 remains separate from the original 17-task FrontierSWE release, and Bug Hunt Bench remains display-only because the result belongs to a full model, effort, and CLI setup. OSWorld 2.0 uses the strict pass rate and records the paired partial score in provenance. Independent LiveBench and Vals evidence now clears the provisional external-consensus gate; the verified rank remains withheld until more weighted benchmark categories clear their coverage thresholds.
Claude Fable 5.1 sits in the Claude Fable family with Claude Fable 5. Its explicit predecessor is Claude Fable 5. 33 of 681 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Agentic at #2, while its lowest eligible position is Reasoning at #4. particularly useful for coding agents, browser research, and computer-use workflows.
Last updated October 17, 2026. Runtime fields remain blank until a sourced snapshot exists.
Questions
How does Claude Fable 5.1 perform overall in AI benchmarks?
Claude Fable 5.1 ranks #4 out of 224 models on the public BenchAlign leaderboard, with a score of 83.18/100. Its evidence status is Supported, and this profile shows 33 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.
Is Claude Fable 5.1 good for knowledge and understanding?
Claude Fable 5.1 ranks #2 out of 184 eligible models for knowledge and understanding, with a public category score of 86.9/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Claude Fable 5.1 good for coding and programming?
Claude Fable 5.1 ranks #3 out of 150 eligible models for coding and programming, with a public category score of 76.8/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Claude Fable 5.1 good for mathematics?
Claude Fable 5.1 has source-displayable benchmark coverage for mathematics, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Claude Fable 5.1 good for reasoning and logic?
Claude Fable 5.1 ranks #4 out of 28 eligible models for reasoning and logic, with a public category score of 85.2/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Claude Fable 5.1 good for agentic tool use and computer tasks?
Claude Fable 5.1 ranks #2 out of 128 eligible models for agentic tool use and computer tasks, with a public category score of 80.9/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Claude Fable 5.1 good for instruction following?
Claude Fable 5.1 has source-displayable benchmark coverage for instruction following, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Which sibling models are related to Claude Fable 5.1?
Claude Fable 5.1 belongs to the Claude Fable family. Related tracked variants include Claude Fable 5. A sibling link indicates shared lineage or a documented configuration relationship; it does not mean the variants have identical pricing, context limits, benchmark evidence, or deployment behavior. Compare before switching.
Does Claude Fable 5.1 have full benchmark coverage on BenchLM?
No. Claude Fable 5.1 currently has 57 source-displayable rows across 681 tracked benchmark slots. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of Claude Fable 5.1?
Claude Fable 5.1 has a documented context window of 1M. That figure is the maximum combined prompt and retained-conversation space reported for this exact model; it is not the maximum output length. The profile keeps output limits separate because providers often publish those limits independently.