Claude Fable 5.1 vs Claude Opus 5.5
Wins 1 of 8 areas
Following instructions
Wins 5 of 8 areas
Coding · Agents · Reasoning · Facts · Math
Claude Opus 5.5 is the stronger all-rounder.Claude Fable 5.1 is better at following instructions.
Both rank among the ten best models we track in reasoning, coding, agents, facts, long documents and math.
Scores updated · 81 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 software06Claude Opus 5.56 of 6 tests
- ReasoningHard problems that need careful thinking04Claude Opus 5.54 of 5 tests · 1 tie
- MathCompetition and research-level math02Claude Opus 5.52 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own12Claude Opus 5.52 of 3 tests
- FactsGetting facts right instead of making them up12Claude Opus 5.52 of 3 tests
- Following instructionsDoing exactly what it is asked10Claude Fable 5.11 of 1 test
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 1 tie
- Long documentsFinding answers in very long texts00Even0 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 5.5 costs 60% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Claude Opus 5.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+14points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+8.7points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+7.6points ahead
Where Claude Fable 5.1 pulls ahead
- Finds the root cause of IT system incidentsAA IT-Bench SRE+11.3points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+7.3points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+1points ahead
Every test, side by side
All 81 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 5.5
- SWE-bench ProClaude Opus 5.5 by 8.781.289.9+8.7
- LMArena · WebDevClaude Opus 5.5 by 68 rating points17461814+68 rating
- LiveBench · Agentic CodingClaude Opus 5.5 by 5.666.171.7+5.6
- SWE-bench MultilingualClaude Opus 5.5 by 4.889.193.9+4.8
- SciCodeClaude Opus 5.5 by 3.863.166.9+3.8
- LiveBench · CodingClaude Opus 5.5 by 2.986.489.3+2.9
AgentsClaude Opus 5.5
- AA IT-Bench SREClaude Fable 5.1 by 11.349.538.2+11.3
- Terminal-Bench 4.0Claude Opus 5.5 by 7.65259.6+7.6
- GDPValClaude Opus 5.5 by 5.662.768.3+5.6
ReasoningClaude Opus 5.5
- Humanity's Last ExamClaude Opus 5.5 by 2.359.161.4+2.3
- CritPtClaude Opus 5.5 by 229.731.7+2
- SimpleBenchClaude Opus 5.5 by 1.886.688.4+1.8
- ARC-AGI-2Claude Opus 5.5 by 1.79091.7+1.7
- LiveBench · Reasoningtie91.792.2tie
FactsClaude Opus 5.5
- AA-Omniscience · Non-hallucinationClaude Opus 5.5 by 1427.441.4+14
- SimpleQA VerifiedClaude Opus 5.5 by 1.470.872.2+1.4
- AA-Omniscience · AccuracyClaude Fable 5.1 by 167.266.2+1
MathClaude Opus 5.5
- FrontierMath Tier 4Claude Opus 5.5 by 7.287.895+7.2
- FrontierMath Tiers 1-3 (v2)Claude Opus 5.5 by 190.291.2+1
- LiveBench · Mathematicstie9797.1tie
Following instructionsClaude Fable 5.1
- LiveBench · Instruction FollowingClaude Fable 5.1 by 7.37365.7+7.3
Other results58 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.
- Science and math (Terminal-Bench-Science 0.1 (AA))Claude Opus 5.5 by 15.743.359+15.7
- Terminal-Bench-Science 0.1Claude Opus 5.5 by 15.743.359+15.7
- BenchCAD w/o toolsClaude Opus 5.5 by 12.460.673+12.4
- AutomationBench (Score)Claude Opus 5.5 by 11.131.442.5+11.1
- GDPval-AA 2.1Claude Opus 5.5 by 111 rating points17351846+111 rating
- Knowledge work (AutomationBench)Claude Opus 5.5 by 11.131.442.5+11.1
- Science and math (Terminal-Bench Science 0.1)Claude Opus 5.5 by 10.752.663.3+10.7
- Terminal-Bench 4.0Claude Opus 5.5 by 10.655.866.4+10.6
- Computer use (OSWorld-2.1)Claude Opus 5.5 by 10.476.887.2+10.4
- PostTrainBenchClaude Opus 5.5 by 9.140.249.3+9.1
- Cybersecurity (CWE-bench v1)Claude Opus 5.5 by 95867+9
- Cybersecurity (CWE-bench)Claude Opus 5.5 by 95867+9
- General capabilities (AA Index v5.0)Claude Opus 5.5 by 94958+9
- HealthBench ProfessionalClaude Fable 5.1 by 8.674.265.6+8.6
- Agentic coding (Terminal-bench 4.0)Claude Opus 5.5 by 8.557.966.4+8.5
- GDPval-AA (Elo)Claude Opus 5.5 by 85 rating points17351820+85 rating
- ArXivMath without toolsClaude Opus 5.5 by 8.382.991.2+8.3
- Agentic coding (Terminal-Bench 4.0 (AA))Claude Opus 5.5 by 7.65259.6+7.6
- OSWorld 2.0 (strict)Claude Opus 5.5 by 741.748.7+7
- DeepSWE 1.1Claude Opus 5.5 by 6.867.474.2+6.8
- SWE-bench MultimodalClaude Opus 5.5 by 6.754.761.4+6.7
- HealthBenchClaude Fable 5.1 by 6.166.760.6+6.1
- AA IntelligenceClaude Opus 5.5 by 4.253.457.6+4.2
- LVBenchClaude Opus 5.5 by 479.783.7+4
- Multimodal understanding (LVBench)Claude Opus 5.5 by 479.783.7+4
- RiemannBenchClaude Opus 5.5 by 465.669.6+4
- Science and math (RiemannBench)Claude Opus 5.5 by 465.669.6+4
- OSWorld 2.0Claude Opus 5.5 by 3.977.981.8+3.9
- OSWorld 2.0 (partial)Claude Opus 5.5 by 3.977.981.8+3.9
- BenchCAD w/ toolsClaude Opus 5.5 by 3.692.696.2+3.6
- Program BenchClaude Opus 5.5 by 3.687.691.2+3.6
- livebench_languageClaude Fable 5.1 by 3.289.586.3+3.2
- AA-OmniscienceClaude Opus 5.5 by 2.943.546.4+2.9
- HealthBench Professional (raw score)Claude Opus 5.5 by 2.974.277.1+2.9
- HLE (with tools)Claude Opus 5.5 by 2.76567.7+2.7
- baby_vision_with_pythonClaude Opus 5.5 by 2.695.197.7+2.6
- HiL-Bench (Tools-allowed)Claude Fable 5.1 by 2.561.559+2.5
- ZeroBench_main w/ python (pass@5)Claude Opus 5.5 by 25961+2
- GraphWalks (256k to 1M, BFS (F1))Claude Opus 5.5 by 1.86566.8+1.8
- HealthBench (raw score)Claude Opus 5.5 by 1.466.768.1+1.4
- mmmu_pro_with_pythonClaude Opus 5.5 by 1.488.489.8+1.4
- OfficeQAClaude Fable 5.1 by 1.380.278.9+1.3
- OfficeQA ProClaude Fable 5.1 by 1.36967.7+1.3
- Knowledge work (Vals Index)Claude Opus 5.5 by 1.265.867+1.2
- Vals IndexClaude Opus 5.5 by 1.265.867+1.2
- charxiv_rq_with_pythonClaude Opus 5.5 by 1.192.793.8+1.1
- OSWorld 2.1Claude Opus 5.5 by 1.180.781.8+1.1
- OSWorld 2.1 partialClaude Opus 5.5 by 1.180.781.8+1.1
- GraphWalks (Up to 128k, BFS (F1))tie91.490.6tie
- PostTrain Bench v1.2tie44.643.8tie
- SWE Marathon v1.1tie45.646.3tie
- ChartoGraphy w/ pythontie88.489tie
- HealthBench length-adjustedtie6060.6tie
- GMMLUtie9494.3tie
- MILUtie9393.1tie
- ARC-AGI-1tie97.597.5tie
- livebench_data_analysistie80.380.3tie
- Toolathlon Verifiedtie77.877.8tie
Questions people ask
Which is better, Claude Fable 5.1 or Claude Opus 5.5?
Claude Opus 5.5 wins five of the eight areas we test: coding, agents, reasoning, facts and math. Claude Fable 5.1 wins following instructions. They are level on images and charts and long documents.
Which is better for coding?
Claude Opus 5.5. It wins 6 of the 6 coding tests both models report; Claude Fable 5.1 wins none.
Which is cheaper?
Claude Fable 5.1 costs $10.00 per million input tokens and $50.00 per million output tokens; Claude Opus 5.5 costs $4.00 and $20.00. That makes Claude Opus 5.5 about 60% cheaper for the same work.
How do you compare the two?
We use the 81 benchmark tests both models have published scores on. The verdict counts the 23 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 58 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.