RankingClaude Fable 5.1

Claude Fable 5.1

Data updated 22 Oct 2026

66 published benchmark measures · 16 benchmark families contribute across 5 task areas. 4 capability estimates available in the full profile. See all results ↓ · Compare published benchmarks →

Model evidence summary

This profile combines published settings. It is not a runnable configuration or a leaderboard rank. Compare measured configurations →

Performance profile

Capabilities

Adjusted comparison score · 0–100

Filled points are supported; hollow points are preliminary. Lines stop at unknown capabilities. Exact values and sources follow below.

Agentic5 families · 0 with independent evidence · Supported97.1

5 core families; 4 direct opponents across 2 labs.

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: 96.3–98.0; 0/13 scenarios unsupported. Without one publisher: 94.0–97.8; 0/15 unsupported. Smoothing check: 90.7–98.9; 0/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage; source sensitivity still requires inspection. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

AutomationBench: 87.5 observed win share
Anthropic, OpenAI · Source 1 Source 2
GDPval: 100.0 observed win share
Anthropic · Source 1
OSWorld: 100.0 observed win share
Anthropic · Source 1 Source 2
OfficeQA: 100.0 observed win share
Anthropic · Source 1 Source 2
Terminal-Bench Science: 75.0 observed win share
OpenAI · Source 1

Hard reasoning4 families · 2 with independent evidence · Supported81.4

4 core families; 61 direct opponents across 10 labs.

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: 75.6–85.1; 0/6 scenarios unsupported. Without one publisher: 71.9–82.7; 0/16 unsupported. Smoothing check: 74.4–84.3; 0/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

GPQA: 40.0 observed win share
OpenAI, artificialanalysis.ai · Source 1 Source 2
ARC-AGI: 52.2 observed win share
Anthropic, OpenAI, arcprize.org · Source 1 Source 2 Source 3 Source 4 Source 5
Humanity’s Last Exam: 100.0 observed win share
Anthropic, OpenAI · Source 1 Source 2 Source 3
FrontierMath: 50.0 observed win share
OpenAI · Source 1

Coding4 families · 1 with independent evidence · Supported93.9

4 core families; 21 direct opponents across 7 labs.

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: 90.8–96.6; 0/6 scenarios unsupported. Without one publisher: 92.6–94.9; 0/19 unsupported. Smoothing check: 86.0–97.0; 0/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts improved on neutral predictions within evaluated coverage. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Terminal-Bench: 88.6 observed win share
Anthropic, OpenAI, tbench.ai · Source 1 Source 2 Source 3
SWE-bench Multilingual: 50.0 observed win share
Anthropic · Source 1
SWE-bench Pro: 100.0 observed win share
Anthropic · Source 1
DeepSWE: 0.0 observed win share
OpenAI · Source 1

Human prefNo comparable evidenceUnknown

0 core families; 0 direct opponents across 0 labs. No admitted core comparisons

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 1/1 scenarios unsupported. Without one publisher: No supported estimate; 1/1 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: cross-family and cross-publisher predictions could not be evaluated from the available comparison network. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Knowledge2 families · 0 with independent evidence · PreliminaryUnknown

2 core families; 2 direct opponents across 1 labs. No connected comparison to the complete reference panel

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 4/4 scenarios unsupported. Without one publisher: No supported estimate; 5/5 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: almost no publisher-held-out comparisons remained estimable. Cross-source validity is unresolved. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Gmmlu: 100.0 observed win share
Anthropic · Source 1
Milu: 100.0 observed win share
Anthropic · Source 1

Multimodal1 families · 0 with independent evidence · Preliminary95.1

1 core families; 2 direct opponents across 1 labs. Needs broader benchmark and opponent support

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 8/8 scenarios unsupported. Without one publisher: No supported estimate; 8/8 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: limited publisher-held-out coverage. Cross-source validity is unresolved. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Chartography: 100.0 observed win share
Anthropic · Source 1 Source 2

Long contextNo comparable evidenceUnknown

0 core families; 0 direct opponents across 0 labs. No admitted core comparisons

Estimated against the same complete panel: gpt-5.6-sol claude-opus-4-8 gemini-3.1-pro . Indirect comparisons assume performance can be summarized by one strength within this capability.

Refit without one family: No supported estimate; 5/5 scenarios unsupported. Without one publisher: No supported estimate; 5/5 unsupported. Smoothing check: No supported estimate; 3/3 unsupported. These are sensitivity checks, not confidence intervals.

Audit: family and publisher holdouts underperformed neutral predictions. Treat this as an exploratory estimate. Validation audit.

Matched family win shares below describe source evidence; they are not averaged to produce the adjusted score.

Scores estimate outcomes against a shared reference panel; they are not accuracy percentages. Sparse or disconnected evidence cannot qualify an overall profile. Open a capability to inspect its evidence.

Reported effort · Mixed settings

Settings reported in this model's published benchmark results, including results outside the aggregate. Effort names are provider-specific. These are not API defaults or equal compute budgets.

Auto thinking: 2 observations
High: 1 observations
Max: 34 observations
Medium: 3 observations
XHigh: 4 observations
Not specified: 65 observations

Mixed settings means multiple settings occur in the evidence. Best across efforts means the source selected its best reported result across settings; it does not mean Max. Unspecified settings stay unknown. This model-summary chart combines reported settings. The leaderboard keeps identified configurations separate and excludes unknown effort.

Inspect each result and its source ↓ · Download effort evidence

Score contributions and missing evidence

16 contributing families across 5 capabilities. Fixed reference panels do not change when the catalog expands.

Capability is fitted jointly across families. Capability estimates below describe different task areas; their weighted sum is not the Capability score.

Results without reviewed compatibility or a reference match remain in the raw evidence below. Coverage counts only contributing results.

Model information & shareable badge

Lab: Anthropic
Catalog status: active
Availability: Public provider catalog; account and region restrictions may apply
Family: Claude Fable
Released: —
Context: 1,000,000 tokens
License: proprietary
Model card: https://platform.claude.com/docs/en/about-claude/model-deprecations
Default Capability family coverage: /badge/claude-fable-5-1.svg

Benchmark scores & sources

Original results, evaluation harnesses, and evidence behind this model.

Benchmark	Bucket	Score	Harness	Evidence	Source-recorded date
Legal Agent Benchmark · 1235-task public subset	Supporting evidence	19.09% Reported settings & source all-pass; five runs; adaptive max; internal bash/Python harness, Sonnet4.6judge;16defective tasks excluded; production safeguards/fallback. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.15.2 · reviewed 2026-10-18	Published configuration Effort: Not specified	lab self-report Supporting evidence outside the reviewed capability core	Reviewed 2026-10-18
Legal Agent Benchmark · 1235-task public subset	Supporting evidence	90.81% Reported settings & source criterion-pass; five runs; adaptive max; internal bash/Python harness, Sonnet4.6judge;16defective tasks excluded; production safeguards/fallback. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.15.2 · reviewed 2026-10-18	Published configuration Effort: Not specified	lab self-report Supporting evidence outside the reviewed capability core	Reviewed 2026-10-18
LMArena Agent Arena	Supporting evidence	14.17	LMArena Agent Arena · max Effort: Not specified	official board Supporting evidence outside the reviewed capability core	2026-10-20
LVBench	Multimodal	79.7% Reported settings & source Comparator maximum available thinking/reasoning when reported, otherwise best available result; exact setting not identified by this grid. All models self-computed without tools. Gemini 1 FPS; Astra 800 frames, Fable 300 frames and Opus 600 frames due to API limits. Unequal frame budgets are excluded from matched comparison. Provider-published launch claim, not independent measurement by Google for every row. All models self-computed without tools. Gemini 1 FPS; Astra 800 frames, Fable 300 frames and Opus 600 frames due to API limits. Unequal frame budgets are excluded from matched comparison. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 18: LVBench / Claude Fable 5.1; methodology Additional Details · reviewed 2026-09-30	Published configuration Effort: Not specified	lab self-report Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.	Reviewed 2026-09-30
MILU	Knowledge	93% Reported settings & source Mean accuracy11languages; adaptive max; five trials; no tools/custom system prompts. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.18.2 · reviewed 2026-10-18	Published configuration Effort: Not specified contributes to capability	lab self-report	Reviewed 2026-10-18
OfficeQA	Agentic	80.2% Reported settings & source Extracted-text Treasury corpus in sandbox; code execution; production Messages API with safeguards/fallback;128koutput cap. Pro133questions. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.15.1 · reviewed 2026-10-18	Published configuration Effort: Not specified contributes to capability	lab self-report	Reviewed 2026-10-18
OfficeQA Pro	Agentic	69% Reported settings & source Extracted-text Treasury corpus in sandbox; code execution; production Messages API with safeguards/fallback;128koutput cap. Pro133questions. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.15.1 · reviewed 2026-10-18	Published configuration Effort: Not specified contributes to capability	lab self-report	Reviewed 2026-10-18
OSWorld · 2.0 August2026 task release	Agentic	77.9% Reported settings & source partial pass@1;108tasks; five runs;1080p;500steps; max effort; Opus4.8grader; task fixes; Fable safety interventions score zero. Same-condition reruns; supersedes earlier OSWorld2results and incompatible with previous task files. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.14.3 · reviewed 2026-10-18	Published configuration Effort: Max contributes to capability	lab self-report	Reviewed 2026-10-18
OSWorld · 2.0 August2026 task release	Agentic	41.7% Reported settings & source strict pass@1;108tasks; five runs;1080p;500steps; max effort; Opus4.8grader; task fixes; Fable safety interventions score zero. Same-condition reruns; supersedes earlier OSWorld2results and incompatible with previous task files. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.14.3 · reviewed 2026-10-18	Published configuration Effort: Max contributes to capability	lab self-report	Reviewed 2026-10-18
PostTrainBench · 1.1	Supporting evidence	40.2% Reported settings & source Comparator maximum available thinking/reasoning when reported, otherwise best available result; exact setting not identified by this grid. All models self-computed; OpenCode harness; 10-hour budget on one NVIDIA H100 GPU. Weighted aggregate across four base models and seven benchmarks. Provider-published launch claim, not independent measurement by Google for every row. All models self-computed; OpenCode harness; 10-hour budget on one NVIDIA H100 GPU. Weighted aggregate across four base models and seven benchmarks. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 9: PostTrainBench 1.1 / Claude Fable 5.1; methodology Additional Details · reviewed 2026-09-30	Published configuration Effort: Not specified	lab self-report Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.	Reviewed 2026-09-30
ProgramBench · 166 golden-task subset	Supporting evidence	87.6% Reported settings & source mini-swe-agent without six-hour timeout; excludes34flaky-reference tasks; tests restricted to reference-passing tests; up to1Mcontext. Hidden-test pass rate, not fraction of completely solved programs. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.11.1 · reviewed 2026-10-18	Published configuration Effort: Not specified	lab self-report Supporting evidence outside the reviewed capability core	Reviewed 2026-10-18
Real-SWE	Supporting evidence	45%	Real-SWE · Fable 5.1 · Claude Code Effort: Not specified	official board Supporting evidence outside the reviewed capability core	2026-10-13
RiemannBench	Supporting evidence	65.6% Reported settings & source Comparator maximum available thinking/reasoning when reported, otherwise best available result; exact setting not identified by this grid. Results sourced from the official Surge public leaderboard. Provider-published launch claim, not independent measurement by Google for every row. Results sourced from the official Surge public leaderboard. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 12: RiemannBench / Claude Fable 5.1; methodology Additional Details · reviewed 2026-09-30	Published configuration Effort: Not specified	lab self-report Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.	Reviewed 2026-09-30
SpireClimb	Supporting evidence	4%	SpireClimb · Slay the Spire 2, Ascension 10 Effort: Not specified	official board Supporting evidence outside the reviewed capability core	2026-09-25
SpireClimb	Supporting evidence	0%	SpireClimb · Slay the Spire 1, Ascension 20 Heart Effort: Not specified	official board Supporting evidence outside the reviewed capability core	2026-09-25
SWE-bench Multilingual	Coding	89.1% Reported settings & source Anthropic reported configuration; adaptive thinking at max effort, default sampling, five-trial mean unless section specifies otherwise; context at most 1M. Comparator configuration follows cited prior card or board. Agent identity is not established as mini-swe-agent. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · Table 8.1.A; section 8.2 · reviewed 2026-10-18	Published configuration Effort: Max contributes to capability	lab self-report	Reviewed 2026-10-18
SWE-bench Multimodal	Supporting evidence	54.7% Reported settings & source Anthropic reported configuration; adaptive thinking at max effort, default sampling, five-trial mean unless section specifies otherwise; context at most 1M. Comparator configuration follows cited prior card or board. Agent identity is not established as mini-swe-agent. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · Table 8.1.A; section 8.2 · reviewed 2026-10-18	Published configuration Effort: Max	lab self-report Supporting evidence outside the reviewed capability core	Reviewed 2026-10-18
SWE-bench Pro	Coding	81.2%	Anthropic SWE-bench Pro reported system Effort: Not specified	lab self-report Legacy self-report lacks a reviewed comparison configuration	2026-09-01
SWE-bench Pro	Coding	81.2% Reported settings & source Anthropic reported configuration; adaptive thinking at max effort, default sampling, five-trial mean unless section specifies otherwise; context at most 1M. Comparator configuration follows cited prior card or board. Agent identity is not established as mini-swe-agent. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · Table 8.1.A; section 8.2 · reviewed 2026-10-18	Published configuration Effort: Max contributes to capability	lab self-report	Reviewed 2026-10-18
Terminal-Bench · 4.0	Coding	55.8% Reported settings & source Claude Code --bare max effort, 15 trials/task over 66 tasks; Anthropic internal reruns. Uses more precise section values rather than rounded headline table. Distinct from2.1; timeouts/resources changed. Fable is the safeguarded deployed configuration; selected tasks may use disclosed Opus fallback, except evaluations explicitly counting safety blocks as failures. Claude Fable 5.1 and Claude Mythos 5.1 System Card · 8.6 · reviewed 2026-10-18	Published configuration Effort: Max contributes to capability	lab self-report	Reviewed 2026-10-18
Terminal-Bench · 4.0	Coding	57.9% Reported settings & source Comparator maximum available thinking/reasoning when reported, otherwise best available result; exact setting not identified by this grid. Argon self-computed; other models from the official public leaderboard. Highest scoring thinking level reported by Terminal-Bench authors. Argon agent identity is not specified in this methodology. Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed; other models from the official public leaderboard. Highest scoring thinking level reported by Terminal-Bench authors. Argon agent identity is not specified in this methodology. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date. Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison. Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 8: Terminal-Bench 4.0 / Claude Fable 5.1; methodology Additional Details · reviewed 2026-09-30	Published configuration Effort: Not specified	lab self-report Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.	Reviewed 2026-09-30
Terminal-Bench · 4.0	Coding	57.9% Reported settings & source Grok 4.7 XHigh (DeepSWE: High); Grok 4.6 High; GPT-5.6 Sol Max; Claude Fable 5.1 Max. Launch comparison; per-model agent harness, evaluation budget and rerun provenance are not specified. Official launch table. Effort follows the column header except the Grok 4.7 DeepSWE asterisk, which explicitly means High. Comparison limit: The launch table does not establish matching agent harnesses and evaluation settings. Retained as a provider claim; not a new comparable evaluation. Grok 4.7 launch comparison · Model Improvements table: Terminal-Bench · reviewed 2026-09-21	Published configuration Effort: Max	lab self-report The launch table does not establish matching agent harnesses and evaluation settings. Retained as a provider claim; not a new comparable evaluation.	Reviewed 2026-09-21
Terminal-Bench 4.0	Coding	57.88%	Terminal-Bench 4.0 reported Effort: Not specified contributes to capability	official board	2026-09-01
Terminal-Bench 4.0 · 4.0	Coding	55.8% Reported settings & source Maximum reported across reasoning efforts; OpenAI research environment or API Provider-published result; comparator measurements are not automatically independently reproduced. GPT-6 Astra: A new generation of intelligence · Coding table / Terminal-Bench 4.0 / Claude Fable 5.1 · reviewed 2026-09-06	Published configuration Effort: Not specified contributes to capability	lab self-report	Reviewed 2026-09-06
Terminal-Bench Science 0.1 · not specified	Agentic	52.6% Reported settings & source Maximum reported across reasoning efforts; OpenAI research environment or API Provider-published result; comparator measurements are not automatically independently reproduced. GPT-6 Astra: A new generation of intelligence · Academic table / Terminal-Bench Science 0.1 / Claude Fable 5.1 · reviewed 2026-09-06	Published configuration Effort: Not specified contributes to capability	lab self-report	Reviewed 2026-09-06