Claude Fable 5.1 vs Claude Opus 5
Wins 7 of 8 areas
Coding · Agents · Reasoning · Facts · Math · Long documents · Following instructions
Wins 0 of 8 areas
—
Claude Fable 5.1 is the stronger all-rounder.Claude Opus 5 is cheaper.
Both rank among the ten best models we track in reasoning, coding, agents and math.
Scores updated · 61 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software50Claude Fable 5.15 of 7 tests · 2 ties
- AgentsCarrying out multi-step tasks on its own30Claude Fable 5.13 of 3 tests
- MathCompetition and research-level math30Claude Fable 5.13 of 3 tests
- ReasoningHard problems that need careful thinking20Claude Fable 5.12 of 6 tests · 4 ties
- FactsGetting facts right instead of making them up21Claude Fable 5.12 of 3 tests
- Long documentsFinding answers in very long texts10Claude Fable 5.11 of 1 test
- Following instructionsDoing exactly what it is asked10Claude Fable 5.11 of 1 test
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 1 tie
Images and charts, long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Claude Opus 5 costs 50% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Fable 5.1 pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+14.6points ahead
- Short factual questions, answered correctlySimpleQA Verified+10.9points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+9.2points ahead
Where Claude Opus 5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+11.8points ahead
Every test, side by side
All 61 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Fable 5.1
- SciCodeClaude Fable 5.1 by 6.763.156.4+6.7
- LMArena · WebDevClaude Fable 5.1 by 52 rating points17411689+52 rating
- LiveBench · CodingClaude Fable 5.1 by 4.986.481.5+4.9
- Terminal-Bench 2.1Claude Fable 5.1 by 2.391.489.1+2.3
- SWE-bench ProClaude Fable 5.1 by 281.279.2+2
- LiveBench · Agentic Codingtie66.165.2tie
- SWE-bench Multilingualtie89.189.5tie
AgentsClaude Fable 5.1
- τ-Bench V3 · BankingClaude Fable 5.1 by 5.147.242.1+5.1
- Terminal-Bench 4.0Claude Fable 5.1 by 35249+3
- GDPValClaude Fable 5.1 by 1.662.761.1+1.6
ReasoningClaude Fable 5.1
- SimpleBenchClaude Fable 5.1 by 686.680.6+6
- Humanity's Last ExamClaude Fable 5.1 by 4.259.154.9+4.2
- CritPttie29.729.1tie
- GPQA Diamondtie93.793.2tie
- LiveBench · Reasoningtie91.791.2tie
- ARC-AGI-2tie9090.4tie
FactsClaude Fable 5.1
- AA-Omniscience · Non-hallucinationClaude Opus 5 by 11.827.439.2+11.8
- SimpleQA VerifiedClaude Fable 5.1 by 10.970.859.9+10.9
- AA-Omniscience · AccuracyClaude Fable 5.1 by 6.367.260.9+6.3
MathClaude Fable 5.1
- FrontierMath Tier 4Claude Fable 5.1 by 14.687.873.2+14.6
- FrontierMath Tiers 1-3 (v2)Claude Fable 5.1 by 4.690.285.6+4.6
- LiveBench · MathematicsClaude Fable 5.1 by 1.39795.7+1.3
Following instructionsClaude Fable 5.1
- LiveBench · Instruction FollowingClaude Fable 5.1 by 9.27363.8+9.2
Other results36 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- Terminal-Bench-Science 0.1Claude Fable 5.1 by 23.652.629+23.6
- OSWorld 2.0Claude Fable 5.1 by 7.377.970.6+7.3
- OSWorld 2.1 partialClaude Fable 5.1 by 6.780.774+6.7
- AA-OmniscienceClaude Fable 5.1 by 6.443.537.1+6.4
- livebench_data_analysisClaude Fable 5.1 by 5.780.374.5+5.7
- SWE-bench MultimodalClaude Opus 5 by 4.754.759.4+4.7
- HiL-Bench (Tools-allowed)Claude Fable 5.1 by 4.561.557+4.5
- LVBenchClaude Fable 5.1 by 4.379.775.4+4.3
- Terminal-Bench 4.0Claude Fable 5.1 by 455.851.8+4
- Artificial Analysis Coding IndexClaude Fable 5.1 by 3.681.678+3.6
- GDPval-AA v2Claude Fable 5.1 by 29 rating points18531824+29 rating
- Toolathlon VerifiedClaude Opus 5 by 2.877.880.6+2.8
- GDPval-AA (Elo)Claude Fable 5.1 by 27 rating points17351708+27 rating
- GDPval-AA 2.1Claude Fable 5.1 by 27 rating points17351708+27 rating
- AA IntelligenceClaude Fable 5.1 by 2.653.450.8+2.6
- OSWorld 2.0 (partial)Claude Fable 5.1 by 2.577.975.4+2.5
- HealthBench ProfessionalClaude Fable 5.1 by 2.362.159.8+2.3
- HealthBench Professional length-adjustedClaude Fable 5.1 by 2.362.159.8+2.3
- HealthBench length-adjustedClaude Fable 5.1 by 2.26057.8+2.2
- Program BenchClaude Fable 5.1 by 2.287.685.4+2.2
- OfficeQAClaude Fable 5.1 by 2.180.278.1+2.1
- OfficeQA ProClaude Fable 5.1 by 2.16966.9+2.1
- OSWorld 2.0 (strict)Claude Fable 5.1 by 2.141.739.6+2.1
- AA Agentic IndexClaude Fable 5.1 by 1.85856.2+1.8
- GMMLUClaude Fable 5.1 by 1.59492.5+1.5
- DeepSWE 1.1Claude Opus 5 by 1.467.468.8+1.4
- HLE (with tools)Claude Fable 5.1 by 1.46563.6+1.4
- GDPval-AA v2 EloClaude Fable 5.1 by 11 rating points18351824+11 rating
- MILUtie9392.1tie
- HealthBench Professional (raw score)tie74.273.4tie
- HealthBench Professional (raw)tie74.273.4tie
- livebench_languagetie89.588.7tie
- GDP (Surge AI)tie85.185.5tie
- HealthBench (raw score)tie66.767.1tie
- HealthBench (raw)tie66.767.1tie
- ARC-AGI-1tie97.597.5tie
Questions people ask
Which is better, Claude Fable 5.1 or Claude Opus 5?
Claude Fable 5.1 wins seven of the eight areas we test: coding, agents, reasoning, facts, math, long documents and following instructions. Claude Opus 5 wins none, but costs 50% less. They are level on images and charts.
Which is better for coding?
Claude Fable 5.1. It wins 5 of the 7 coding tests both models report; Claude Opus 5 wins none, and 2 are ties.
Which is cheaper?
Claude Fable 5.1 costs $10.00 per million input tokens and $50.00 per million output tokens; Claude Opus 5 costs $5.00 and $25.00. That makes Claude Opus 5 about 50% cheaper for the same work.
How do you compare the two?
We use the 61 benchmark tests both models have published scores on. The verdict counts the 25 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 36 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.