Claude Fable 5 vs Claude Opus 5
Wins 4 of 8 areas
Facts · Math · Long documents · Following instructions
Wins 3 of 8 areas
Coding · Agents · Reasoning
Claude Fable 5 wins more areas, narrowly.Claude Opus 5 is cheaper and better at agents.
Both rank among the ten best models we track in reasoning, coding, agents and math.
Scores updated · 62 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.
- MathCompetition and research-level math20Claude Fable 52 of 3 tests · 1 tie
- FactsGetting facts right instead of making them up21Claude Fable 52 of 3 tests
- Long documentsFinding answers in very long texts10Claude Fable 51 of 1 test
- Following instructionsDoing exactly what it is asked10Claude Fable 51 of 1 test
- AgentsCarrying out multi-step tasks on its own15Claude Opus 55 of 6 tests
- CodingWriting and fixing software25Claude Opus 55 of 8 tests · 1 tie
- ReasoningHard problems that need careful thinking12Claude Opus 52 of 6 tests · 3 ties
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 2 ties
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 three tests each model wins by the widest margin. Scores are out of 100.
Where Claude Fable 5 pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+17points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+12points ahead
- Short factual questions, answered correctlySimpleQA Verified+10.8points ahead
Where Claude Opus 5 pulls ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+6.6points ahead
- Real work tasks from 44 professionsGDPVal+5.6points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+4.5points ahead
Every test, side by side
All 62 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 5
- LMArena · WebDevClaude Opus 5 by 68 rating points16241692+68 rating
- SciCodeClaude Fable 5 by 4.66156.4+4.6
- LiveBench · CodingClaude Fable 5 by 4.58681.5+4.5
- Terminal-Bench 2.1Claude Opus 5 by 4.584.689.1+4.5
- LiveBench · Agentic CodingClaude Opus 5 by 362.265.2+3
- SWE-bench MultilingualClaude Opus 5 by 2.986.689.5+2.9
- SWE-bench VerifiedClaude Opus 5 by 19596+1
- SWE-bench Protie8079.2tie
AgentsClaude Opus 5
- Terminal-Bench 4.0Claude Opus 5 by 6.642.449+6.6
- GDPValClaude Opus 5 by 5.655.661.2+5.6
- τ-Bench V3 · BankingClaude Opus 5 by 438.142.1+4
- BrowseCompClaude Opus 5 by 3.487.490.8+3.4
- OSWorld-VerifiedClaude Fable 5 by 1.68583.4+1.6
- MCP AtlasClaude Opus 5 by 1.184.785.8+1.1
ReasoningClaude Opus 5
- LiveBench · ReasoningClaude Opus 5 by 1.589.791.2+1.5
- SimpleBenchClaude Fable 5 by 1.381.980.6+1.3
- ARC-AGI-2Claude Opus 5 by 1.289.290.4+1.2
- GPQA Diamondtie92.693.2tie
- Humanity's Last Examtie55.554.9tie
- CritPttie28.629.1tie
FactsClaude Fable 5
- SimpleQA VerifiedClaude Fable 5 by 10.870.759.9+10.8
- AA-Omniscience · AccuracyClaude Fable 5 by 4.565.360.9+4.5
- AA-Omniscience · Non-hallucinationClaude Opus 5 by 2.836.439.2+2.8
Images and chartsEven
- MMMU-Protie84.284.7tie
- LMArena · Visiontie13251321tie
MathClaude Fable 5
- FrontierMath Tier 4Claude Fable 5 by 1790.273.2+17
- FrontierMath Tiers 1-3 (v2)Claude Fable 5 by 1.48785.6+1.4
- LiveBench · Mathematicstie9695.7tie
Following instructionsClaude Fable 5
- LiveBench · Instruction FollowingClaude Fable 5 by 1275.863.8+12
Other results32 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.
- GDP (Surge AI)Claude Opus 5 by 55.729.885.5+55.7
- GDPval-AA (Elo)Claude Fable 5 by 224 rating points19321708+224 rating
- GDPval-AA 2.1Claude Opus 5 by 113 rating points15951708+113 rating
- GDPval-AA v2Claude Opus 5 by 101 rating points17231824+101 rating
- GDPval-AA v2 EloClaude Opus 5 by 101 rating points17231824+101 rating
- OfficeQA ProClaude Opus 5 by 957.966.9+9
- Terminal-Bench-Science 0.1Claude Opus 5 by 8.621.430+8.6
- JobBenchClaude Opus 5 by 8.357.465.7+8.3
- Terminal-Bench 4.0Claude Opus 5 by 7.344.551.8+7.3
- AA-OmniscienceClaude Fable 5 by 6.243.337.1+6.2
- HealthBench ProfessionalClaude Fable 5 by 6.26659.8+6.2
- Agents' Last ExamClaude Opus 5 by 5.925.731.6+5.9
- HealthBench (raw)Claude Opus 5 by 5.961.267.1+5.9
- livebench_data_analysisClaude Fable 5 by 5.980.574.5+5.9
- SWE-bench MultimodalClaude Opus 5 by 5.354.159.4+5.3
- AA Agentic IndexClaude Opus 5 by 5.25156.2+5.2
- HealthBench Professional (raw)Claude Opus 5 by 4.568.973.4+4.5
- OSWorld 2.0Claude Opus 5 by 4.566.170.6+4.5
- OSWorld 2.0 (strict)Claude Opus 5 by 3.536.139.6+3.5
- Toolathlon VerifiedClaude Opus 5 by 2.777.980.6+2.7
- OSWorld 2.0 (partial)Claude Opus 5 by 2.572.975.4+2.5
- livebench_languageClaude Fable 5 by 290.788.7+2
- Legal Agent BenchmarkClaude Fable 5 by 1.613.311.7+1.6
- Artificial Analysis Coding IndexClaude Opus 5 by 1.576.578+1.5
- AA IntelligenceClaude Opus 5 by 1.249.650.8+1.2
- DeepSWE 1.1Claude Fable 5 by 1.27068.8+1.2
- GMMLUClaude Fable 5 by 1.193.692.5+1.1
- ARC-AGI-1Claude Fable 5 by 198.597.5+1
- Program Benchtie86.385.4tie
- MILUtie92.992.1tie
- HiL-Bench (Tools-allowed)tie56.357tie
- HLE (with tools)tie63.863.6tie
Questions people ask
Which is better, Claude Fable 5 or Claude Opus 5?
Claude Fable 5 wins four of the eight areas we test: facts, math, long documents and following instructions. Claude Opus 5 wins coding, agents and reasoning, and costs 50% less. They are level on images and charts.
Which is better for coding?
Claude Opus 5. It wins 5 of the 8 coding tests both models report; Claude Fable 5 wins 2, and 1 is a tie.
Which is cheaper?
Claude Fable 5 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 62 benchmark tests both models have published scores on. The verdict counts the 30 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 32 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.