BenchLeader
Menu
Best LLMsChoose a modelLeaderboardsModelsBenchmarksProvidersReleases12

Claude Fable 5 vs Gemini 3.1 Pro

Full profiles: Claude Fable 5 (Anthropic) and Gemini 3.1 Pro (Google): every benchmark, price and rank.

Updated · Data from 14 sources

As of 24 Oct 2026, Claude Fable 5 (max) leads on quality (69.3 vs 60.6 on the BenchLeader Index), but Gemini 3.1 Pro costs 4.4× less, so it is the better buy unless you need the extra quality. Pick Claude Fable 5 for knowledge (81 vs 64); pick Gemini 3.1 Pro for instruction following (68 vs 63).

Claude Fable 5 benchmarks max Anthropic Gemini 3.1 Pro benchmarks GoogleOpen in the comparison tool

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingClaude Fable 574 vs 57 category score; SciCode 61.0% vs 58.7%High
Agents & tool useClaude Fable 565 vs 57 category score; τ³-Banking (AA) 38.1% vs 21.4%High
MathsClaude Fable 572 vs 61 category score; FrontierMath Tiers 1–3 87.0% vs 59.6%High
ReasoningClaude Fable 574 vs 68 category score; SimpleBench 81.9% vs 79.6%Medium
KnowledgeClaude Fable 581 vs 64 category score; AA-Omniscience 43.3 vs 31.9High
Long documentsEither66 vs 66 category score; AA-LCR 82.3% vs 82.0%Either
Following instructionsGemini 3.1 Pro68 vs 63 category score; IFBench 77.1% vs 63.5%Medium
Chat quality (human preference)–Only Gemini 3.1 Pro has an Arena rating (59 category score)Not comparable
Lowest costGemini 3.1 Pro$4.50 vs $20.00 per 1M blended; agentic coding session $0.168 vs $0.800–
Fastest first answerGemini 3.1 Pro21.1 s (Gemini 3.1 Pro) vs 59.8 s (Claude Fable 5 (max)) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • Claude Fable 5 (max) leads on quality: 69.3 vs 60.6.
  • Claude Fable 5 (max) is stronger in agents & tools, coding, composite, knowledge, long context, maths, reasoning.
  • Gemini 3.1 Pro is stronger in instruction following, human preference, multimodal.
  • Gemini 3.1 Pro is 4.4× cheaper ($4.50 vs $20.00 per 1M blended).
  • Gemini 3.1 Pro streams 1.9× faster (132 vs 70 tokens per second).

At a glance

SpecClaude Fable 5Gemini 3.1 Pro
DeveloperAnthropicGoogle
Released9 Jun 202619 Feb 2026
Context window (tokens)1,000,0001,048,576
Compared atClaude Fable 5 (max)Gemini 3.1 Pro
Open weightsNoNo
Price in / out / cached input per 1M$10.00 / $50.00 / $1.00$2.00 / $12.00 / $0.200

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryClaude Fable 5 (max)Gemini 3.1 ProLeader
Reasoning73.868.2Claude Fable 5
Coding73.857.0Claude Fable 5
Agents & tools65.057.1Claude Fable 5
Maths72.460.8Claude Fable 5
Knowledge81.063.9Claude Fable 5
Instruction following63.167.5Gemini 3.1 Pro
Human preference–59.1not comparable
Multimodal–64.8not comparable
Long context66.366.1level
Composite85.049.2Claude Fable 5

Where sources disagree

Across the 36 benchmarks both have, Claude Fable 5 leads 31, Gemini 3.1 Pro leads 5. Every overall index that rates both puts Claude Fable 5 ahead.

Overall indexClaude Fable 5 (max)Gemini 3.1 ProLeader
BenchLeader Index69.360.6Claude Fable 5
AA Intelligence Index v4.3.249.629.7Claude Fable 5
Arena Text1503†1487Claude Fable 5

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.

MeasureClaude Fable 5 (max)Gemini 3.1 Pro
Blended price per 1M tokens$20.00$4.50
Input / output per 1M$10.00 / $50.00$2.00 / $12.00
Cached input per 1M$1.00$0.200
Price per task: chat reply$0.019$0.0044
Price per task: summarize a 30-page report$0.150$0.031
Price per task: code edit$0.135$0.030
Price per task: agentic coding session$0.800$0.168
Price per task: structured extraction$0.030$0.0064
Cost to run a full benchmark suite$8.75$1.30
Output speed70 tok/s132 tok/s
Time to first answer59.8 s21.1 s
Time to a complete answer67.0 s24.9 s

Per token is not per task: on Artificial Analysis' full suite Claude Fable 5 (max) cost $8.75 and Gemini 3.1 Pro $1.30, because Claude Fable 5 (max) used about 1.5× as many tokens for the same work, which a per-token price hides.

Each model is compared at its best-ranked reasoning-effort setting. Claude Fable 5 (max) scores 69.3 and takes 60 s to its first answer token; Claude Fable 5 scores 67.1 in 4 s, Claude Fable 5 (high) scores 64.6 in 4 s. Gemini 3.1 Pro scores 60.6 and takes 21 s to its first answer token; Gemini 3.1 Pro (high) scores 60.3 in 21 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

164 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherClaude Fable 5 (max)Gemini 3.1 Pro
BenchLeader IndexBenchLeader69.360.6
Agents & tools scoreBenchLeader65.057.1
Coding scoreBenchLeader73.857.0
Composite scoreBenchLeader85.049.2
Instruction following scoreBenchLeader63.167.5
Knowledge scoreBenchLeader81.063.9
Long context scoreBenchLeader66.366.1
Maths scoreBenchLeader72.460.8
Reasoning scoreBenchLeader73.868.2
Human preference scoreBenchLeader–59.1
Multimodal scoreBenchLeader–64.8
Blended price $/MBenchLeader$20.00$4.50
Output speedBenchLeader70 tok/s132 tok/s
Time to first answerBenchLeader59.8 s21.1 s
Context windowBenchLeader1,000,000 tokens1,048,576 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub85.9%94.1%
Humanity's Last ExamScale AI / CAIS–46.4%
SimpleBenchSimpleBench81.9%79.6%
Arena Hard PromptsArena (LMArena)1529†1508
LiveBench 2.0 ReasoningLiveBench89.7%†84.0%
LiveBench 1.0 ReasoningLiveBench89.7%84.0%†
GPQA Diamond (AA)Artificial Analysis92.6%94.1%
Humanity's Last Exam (AA)Artificial Analysis55.5%47.0%
GPQA Diamond (Vals)Vals AI93.2%†95.5%†
EnigmaEvalScale AI SEAL39.3%†36.8%†
Kagi LLM BenchmarkKagi LLM Benchmark88.8%†–
ARC-AGI-1ARC Prize98.5%98.0%
ARC-AGI-2ARC Prize89.2%77.1%
ARC-AGI-3ARC Prize–0.4%
CritPtArtificial Analysis28.6%17.7%
MysteryMechanismVals AI–20.7%†
Chess PuzzlesEpoch AI Benchmarking Hub47.0%51.0%†
Mystery Game PuzzlesEpoch AI Benchmarking Hub52.0%34.0%†
CL-benchEpoch AI Benchmarking Hub–20.8%
CL-bench LifeEpoch AI Benchmarking Hub–16.9%
LMCAEpoch AI Benchmarking Hub60.3%53.8%†
DTBenchEpoch AI Benchmarking Hub98.4%95.5%†
ForecastBenchForecastBench66.0%64.3%
BTF-3Epoch AI Benchmarking Hub13.0%†–
Coding
SciCodeSciCode61.0%58.7%
Show all 164 benchmarks
MetricPublisherClaude Fable 5 (max)Gemini 3.1 Pro
Coding
WeirdMLWeirdML91.9%72.1%
FrontierCodeCognition53.5%†–
GSO-BenchGSO-Bench78.4%†22.6%
Arena CodingArena (LMArena)1551†1522
Arena WebDevArena (LMArena)1624†1447
LiveBench 2.0 CodingLiveBench98.3%†66.7%
LiveBench 1.0 CodingLiveBench86.0%76.5%†
LiveCodeBenchVals AI89.8%†88.5%†
IOIVals AI–51.8%†
SWE-bench (Vals)Vals AI95.0%†78.8%†
SWE-Bench ProScale AI SEAL–46.1%†
Code MigrationVals AI55.1%†17.3%†
ProgramBenchVals AI2.0%†0.0%†
Vibe Code Bench 1-100Vals AI–6.7%†
Vibe Code Bench v1.1Vals AI90.3%†32.0%†
SWE-Bench Pro (private)Scale AI SEAL–46.1%†
SWE Atlas: Codebase QnAScale AI SEAL39.0%†13.5%
SWE Atlas: RefactoringScale AI SEAL54.8%†33.8%
SWE Atlas: Test WritingScale AI SEAL55.6%†29.8%
MirrorCodeEpoch AI Benchmarking Hub63.9%†8.9%†
Surface Evolver BenchEpoch AI Benchmarking Hub95.0%†–
DeepSWE v1.1Epoch AI Benchmarking Hub69.7%11.7%
ALE-BenchEpoch AI Benchmarking Hub2041.3†1160.6
AlgoTuneAlgoTune–2.0
FrontierSWEEpoch AI Benchmarking Hub47.0%–
Terminal-Bench 4.0 (AA)Artificial Analysis42.4%4.0%
Terminal-Bench 2.1 (AA)Artificial Analysis84.6%73.8%
Startrise FrontendStartrise73.8%†–
Agents & tools
Terminal-BenchTerminal-Bench44.5%80.2%
Remote Labor IndexScale AI / CAIS15.8%†–
APEX-AgentsMercor63.6%†35.1%
Arena AgentArena (LMArena)8.3†-3.7
LiveBench Long AgenticLiveBench53.2%†23.9%
LiveBench BrowserLiveBench92.0%†4.0%
LiveBench Agentic CodingLiveBench62.2%44.1%†
Terminal-Bench HardArtificial Analysis62.9%53.8%
τ²-Bench Telecom (AA)Artificial Analysis98.5%95.6%
Terminal-Bench 2.1 (Vals)Vals AI80.5%†70.8%†
MCP AtlasScale AI SEAL83.3%†78.2%†
HiL-BenchScale AI SEAL56.3%†35.3%
τ³-bench Bankingτ²-bench39.7%26.0%†
GDPval-AA v2.1Artificial Analysis–14.6%
τ³-Banking (AA)Artificial Analysis38.1%21.4%
ITBench SRE (AA)Artificial Analysis–30.3%
Analyst Agent (AA)Artificial Analysis48.8%41.3%
APEX-Agents (AA)Artificial Analysis–32.0%
Finance Agent v2Vals AI56.3%†43.0%†
Harvey's Legal Agent Benchmark v1Vals AI11.3%†0.0%†
Legal Research BenchVals AI49.5%†20.7%†
SREBenchVals AI–0.0%†
Tax Agent BenchVals AI38.3%†9.3%†
Terminal-Bench 2.0 (Vals)Vals AI–67.4%†
Terminal-Bench 4.0 (Vals)Vals AI41.4%†2.5%†
Terminal-Bench ScienceVals AI15.7%†1.4%†
Web Search IndexVals AI46.9%†–
Arena SearchArena (LMArena)1229†–
BALROGBALROG–57.0%
DeepResearch BenchDeepResearch Bench–47.8%†
PostTrainBenchEpoch AI Benchmarking Hub41.8%22.0%
ExploitBenchEpoch AI Benchmarking Hub–26.1%
METR Time HorizonsMETR–77.0%
GBAEvalEpoch AI Benchmarking Hub74.5%†0.8%
Vending-Bench 2Andon Labs4966.6911.2
AutomationBenchArtificial Analysis54.1%35.4%
GDP.pdf (AA)Artificial Analysis24.0%17.8%
Harvey LAB-AA v1.1Harvey–0.0%
EnterpriseOps-GymArtificial Analysis51.1%42.2%
AA-Briefcase v1.1Artificial Analysis1537456
BrowseCompOpenAI–31.2%
DeepSearchQAGoogle–60.2%
FACTS SearchGoogle–83.7%
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub87.0%59.6%
FrontierMath Tier 4Epoch AI Benchmarking Hub90.2%26.8%
OTIS Mock AIMEEpoch AI Benchmarking Hub99.7%95.6%
ProofBenchVals AI95.0%26.0%
LiveBench MathematicsLiveBench96.0%91.0%†
AIME (Vals)Vals AI–98.1%†
AIME 2026MathArena–98.3%
HMMT February 2026MathArena–94.7%
USAMO 2026MathArena–74.4%
MathArena ApexMathArena–60.9%
Arena MathsArena (LMArena)1521†1489
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub70.7%†73.5%†
LiveBench 2.0 Data AnalysisLiveBench80.5%†78.5%
LiveBench ResearchLiveBench86.9%†65.7%
LiveBench 1.0 Data AnalysisLiveBench80.5%78.5%†
AA-OmniscienceArtificial Analysis43.331.9
MMLU-ProVals AI91.5%†91.0%†
LegalBenchVals AI88.6%†87.4%†
CorpFinVals AI71.8%†64.5%†
TaxEvalVals AI76.9%†72.9%†
MedQAVals AI–96.4%†
PRBench FinanceScale AI SEAL53.9%†41.9%
PRBench LegalScale AI SEAL52.6%†44.0%
MultiNRCScale AI SEAL–64.7%
BioMysteryBenchVals AI–60.0%†
CaseLaw v2Vals AI–64.8%†
Excel Modeling BenchmarkVals AI73.7%†52.6%†
MedCodeVals AI56.1%†59.1%†
MedScribeVals AI88.5%†76.1%†
Public Benefits Bench v1.1Vals AI70.4%†53.8%†
DrugDiscoveryBenchScale AI SEAL–41.9%†
EBR-benchEpoch AI Benchmarking Hub39.5%14.3%
GDP.pdfEpoch AI Benchmarking Hub29.8%17.0%
AA-Omniscience: accuracyArtificial Analysis65.3%54.9%
AA-Omniscience: non-hallucinationArtificial Analysis36.4%49.1%
Global-MMLU-Lite (AA)Artificial Analysis–93.2%
Instruction following
LiveBench LanguageLiveBench90.7%85.4%†
LiveBench Instruction FollowingLiveBench75.8%79.1%†
IFBenchArtificial Analysis63.5%77.1%
MultiChallengeScale AI SEAL–71.4%
TutorBenchScale AI SEAL–53.0%
Arena Instruction FollowingArena (LMArena)1508†1480
Human preference
Arena TextArena (LMArena)1503†1487
EQ-Bench 4EQ-Bench1340†1142
Arena Creative WritingArena (LMArena)1501†1481
Arena Multi-turnArena (LMArena)1517†1496
Arena Longer QueriesArena (LMArena)1522†1499
Multimodal
Arena VisionArena (LMArena)1325†1296
LiveBench DesignLiveBench84.7%†61.6%
MMMU-ProArtificial Analysis–82.4%
MMMU-Pro (official)MMMU–80.5%†
MMMU-Pro (Vals)Vals AI89.3%†88.2%†
MortgageTaxVals AI68.9%†69.4%†
SAGEVals AI51.9%†48.7%†
VTBScale AI SEAL–29.0%
Blueprint-Bench 2Epoch AI Benchmarking Hub38.6%†26.5%
Long context
AA-LCRArtificial Analysis82.3%82.0%
Arena DocumentArena (LMArena)1499†1459
MLCRArtificial Analysis20.6%11.1%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub163.4†155.3
LiveBench 2.0LiveBench83.6%†54.9%
LiveBench 1.0LiveBench83.0%77.0%†
AA Intelligence Index v4.3.2Artificial Analysis49.629.7
Vals RSI IndexVals AI24.3%–
Vals Science IndexVals AI–27.4%†
Vals Multimodal IndexVals AI74.2%†56.1%†
Safety & honesty
FORTRESSScale AI SEAL–29.8%
MASKScale AI SEAL–42.4%

Data as of 2026-10-24. Best configuration of each model; every score links to its source on the model pages.

What changed

  • Claude Fable 5 (max): 1 new result (Vals RSI Index).
  • Claude Fable 5 (max): 1 score revision.
  • Gemini 3.1 Pro: 2 score revisions.
  • Claude Fable 5 (max): 1 new result (ForecastBench).
  • Gemini 3.1 Pro: 1 score revision.
  • Gemini 3.1 Pro: BenchLeader Index 62.4 → 60.7.
  • Gemini 3.1 Pro: 8 new results (LiveBench 2.0, LiveBench Long Agentic, LiveBench 2.0 Reasoning…).
  • Claude Fable 5 (max): 4 new results (GDPval-AA v2.1, AA-Omniscience: accuracy, AA-Omniscience: non-hallucination…).
  • Gemini 3.1 Pro: 1 new result (USAMO 2026).
  • Gemini 3.1 Pro: 1 score revision.

Rankings: Best LLM overall · Best LLM for coding · Best LLM for agents · Best cheap LLMs

Claude Fable 5 vs Gemini 3.1 Pro: questions

Is Claude Fable 5 better than Gemini 3.1 Pro?
Claude Fable 5 (max) leads on quality: 69.3 vs 60.6. The BenchLeader Index combines every independent quality benchmark; Claude Fable 5 (max) is ahead overall as of 2026-10-24, but check the category scores for your use.
Which is cheaper, Claude Fable 5 or Gemini 3.1 Pro, and will that change?
Gemini 3.1 Pro is cheaper today: $4.50 against $20.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Claude Fable 5 better than Gemini 3.1 Pro for coding?
Claude Fable 5 scores higher in coding (74 vs 57 on the category index, where 50 is average). On SciCode, the top shared coding benchmark, it is 61.0% for Claude Fable 5 against 58.7% for Gemini 3.1 Pro.
Is Claude Fable 5 better than Gemini 3.1 Pro for agents and tool use?
Claude Fable 5 scores higher in agentic tasks (65 vs 57 on the category index, where 50 is average). On Terminal-Bench, the top shared agents & tools benchmark, it is 44.5% for Claude Fable 5 against 80.2% for Gemini 3.1 Pro.
Which is better for long documents?
They are level on long-context benchmarks (66 vs 66). Both take about 1M tokens: Claude Fable 5 1,000,000, Gemini 3.1 Pro 1,048,576.
Which has lower latency?
Gemini 3.1 Pro answers first: 21.1 s against 59.8 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
Which is faster, Claude Fable 5 or Gemini 3.1 Pro?
Gemini 3.1 Pro streams faster: 132 against 70 output tokens per second.
Do independent benchmarks agree on Claude Fable 5 vs Gemini 3.1 Pro?
Across the 36 benchmarks both have, Claude Fable 5 leads 31, Gemini 3.1 Pro leads 5. Every overall index that rates both puts Claude Fable 5 ahead.
When was Claude Fable 5 released?
Claude Fable 5 was released on 9 Jun 2026.