Claude Fable 5 vs Claude Sonnet 4.6
Wins 8 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Long documents · Following instructions
Wins 0 of 8 areas
—
Claude Fable 5 is the stronger all-rounder.Claude Sonnet 4.6 is cheaper.
Scores updated · 52 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 software80Claude Fable 58 of 8 tests
- AgentsCarrying out multi-step tasks on its own70Claude Fable 57 of 7 tests
- ReasoningHard problems that need careful thinking50Claude Fable 55 of 5 tests
- Images and chartsUnderstanding pictures, charts and video20Claude Fable 52 of 2 tests
- Following instructionsDoing exactly what it is asked20Claude Fable 52 of 2 tests
- FactsGetting facts right instead of making them up21Claude Fable 52 of 3 tests
- MathCompetition and research-level math10Claude Fable 51 of 1 test
- Long documentsFinding answers in very long texts10Claude Fable 51 of 1 test
Math and long documents 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 Sonnet 4.6 costs 70% 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 pulls ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+39.4points ahead
- Short factual questions, answered correctlySimpleQA Verified+37.9points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+30.9points ahead
Where Claude Sonnet 4.6 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+15.2points ahead
Every test, side by side
All 52 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Fable 5
- SWE-bench ProClaude Fable 5 by 21.98058.1+21.9
- LiveBench · Agentic CodingClaude Fable 5 by 19.662.242.6+19.6
- SWE-bench VerifiedClaude Fable 5 by 15.49579.6+15.4
- Terminal-Bench 2.1Claude Fable 5 by 13.484.671.2+13.4
- SciCodeClaude Fable 5 by 10.96150.1+10.9
- LMArena · WebDevClaude Fable 5 by 102 rating points16241522+102 rating
- Terminal-Bench HardClaude Fable 5 by 9.962.953+9.9
- LiveBench · CodingClaude Fable 5 by 6.78679.3+6.7
AgentsClaude Fable 5
- Terminal-Bench 4.0Claude Fable 5 by 39.442.43+39.4
- MCP AtlasClaude Fable 5 by 23.484.761.3+23.4
- GDPValClaude Fable 5 by 18.955.636.7+18.9
- AA ApexAgentsClaude Fable 5 by 15.343.328+15.3
- BrowseCompClaude Fable 5 by 12.787.474.7+12.7
- OSWorld-VerifiedClaude Fable 5 by 12.58572.5+12.5
- τ-Bench V3 · BankingClaude Fable 5 by 3.738.134.4+3.7
ReasoningClaude Fable 5
- ARC-AGI-2Claude Fable 5 by 30.989.258.3+30.9
- CritPtClaude Fable 5 by 25.528.63.1+25.5
- Humanity's Last ExamClaude Fable 5 by 21.955.533.6+21.9
- GPQA DiamondClaude Fable 5 by 5.192.687.5+5.1
- LiveBench · ReasoningClaude Fable 5 by 4.989.784.8+4.9
FactsClaude Fable 5
- SimpleQA VerifiedClaude Fable 5 by 37.970.732.8+37.9
- AA-Omniscience · AccuracyClaude Fable 5 by 24.565.340.9+24.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.6 by 15.236.451.6+15.2
Images and chartsClaude Fable 5
- MMMU-ProClaude Fable 5 by 10.984.273.3+10.9
- LMArena · VisionClaude Fable 5 by 42 rating points13251283+42 rating
Following instructionsClaude Fable 5
- LiveBench · Instruction FollowingClaude Fable 5 by 12.675.863.2+12.6
- IFBenchClaude Fable 5 by 6.963.556.6+6.9
Other results23 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 Sonnet 4.6 by 48.829.878.6+48.8
- DeepSWE 1.1Claude Fable 5 by 407030+40
- GDPval-AA v2Claude Fable 5 by 328 rating points17231395+328 rating
- Blueprint-Bench 2Claude Fable 5 by 31.938.66.7+31.9
- AA-OmniscienceClaude Fable 5 by 31.143.312.2+31.1
- GDPval-AA (Elo)Claude Fable 5 by 256 rating points19321676+256 rating
- τ²-Bench Telecom (AA run)Claude Fable 5 by 22.898.575.7+22.8
- HealthBench ProfessionalClaude Fable 5 by 21.86644.2+21.8
- ToolathlonClaude Fable 5 by 20.761.741+20.7
- AA IntelligenceClaude Fable 5 by 19.549.630.1+19.5
- AA Agentic IndexClaude Fable 5 by 17.95133.1+17.9
- Toolathlon Pass@∞Claude Fable 5 by 17.655.638+17.6
- HLE (with tools)Claude Fable 5 by 1763.846.8+17
- livebench_languageClaude Fable 5 by 14.690.776.1+14.6
- Artificial Analysis Coding IndexClaude Fable 5 by 13.576.563+13.5
- ARC-AGI-1Claude Fable 5 by 12.598.586+12.5
- Legal Agent BenchmarkClaude Fable 5 by 7.913.35.4+7.9
- Legal Agent Benchmark (Harvey's Held-Out Set)Claude Fable 5 by 7.913.35.4+7.9
- Finance Agent v2Claude Fable 5 by 5.356.351+5.3
- OfficeQA ProClaude Fable 5 by 4.557.953.4+4.5
- Toolathlon Avg turnsClaude Fable 5 by 3.319.816.5+3.3
- LiveBenchClaude Fable 5 by 2.878.375.5+2.8
- livebench_data_analysisClaude Fable 5 by 2.580.578+2.5
Questions people ask
Which is better, Claude Fable 5 or Claude Sonnet 4.6?
Claude Fable 5 wins all eight areas we test: coding, agents, reasoning, facts, images and charts, math, long documents and following instructions. Claude Sonnet 4.6 wins none, but costs 70% less.
Which is better for coding?
Claude Fable 5. It wins 8 of the 8 coding tests both models report; Claude Sonnet 4.6 wins none.
Which is cheaper?
Claude Fable 5 costs $10.00 per million input tokens and $50.00 per million output tokens; Claude Sonnet 4.6 costs $3.00 and $15.00. That makes Claude Sonnet 4.6 about 70% cheaper for the same work.
How do you compare the two?
We use the 52 benchmark tests both models have published scores on. The verdict counts the 29 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 23 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.