Claude Fable 5 vs MiniMax M3
Wins 5 of 8 areas
Coding · Agents · Reasoning · Images and charts · Math
Wins 0 of 8 areas
—
Claude Fable 5 is the stronger all-rounder.MiniMax M3 is cheaper.
Scores updated · 43 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 thinking40Claude Fable 54 of 5 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video20Claude Fable 52 of 2 tests
- MathCompetition and research-level math10Claude Fable 51 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
- Following instructionsDoing exactly what it is asked11Even1 each
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.
MiniMax M3 costs 98% 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
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+48.7points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+40.4points ahead
- Common-sense trick questionsSimpleBench+36.1points ahead
Where MiniMax M3 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+45.2points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+19.4points ahead
Every test, side by side
All 43 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Fable 5
- LiveBench · Agentic CodingClaude Fable 5 by 21.562.240.7+21.5
- SWE-bench ProClaude Fable 5 by 218059+21
- Terminal-Bench HardClaude Fable 5 by 20.562.942.4+20.5
- Terminal-Bench 2.1Claude Fable 5 by 19.484.665.2+19.4
- LiveBench · CodingClaude Fable 5 by 17.88668.2+17.8
- SWE-bench VerifiedClaude Fable 5 by 14.59580.5+14.5
- LMArena · WebDevClaude Fable 5 by 143 rating points16251482+143 rating
- SciCodeClaude Fable 5 by 13.96147.1+13.9
AgentsClaude Fable 5
- Terminal-Bench 4.0Claude Fable 5 by 40.442.42+40.4
- τ-Bench V3 · BankingClaude Fable 5 by 22.838.115.3+22.8
- GDPValClaude Fable 5 by 18.355.637.3+18.3
- AA ApexAgentsClaude Fable 5 by 15.643.327.7+15.6
- OSWorld-VerifiedClaude Fable 5 by 14.98570.1+14.9
- MCP AtlasClaude Fable 5 by 10.584.774.2+10.5
- BrowseCompClaude Fable 5 by 3.987.483.5+3.9
ReasoningClaude Fable 5
- SimpleBenchClaude Fable 5 by 36.181.945.8+36.1
- CritPtClaude Fable 5 by 24.928.63.7+24.9
- Humanity's Last ExamClaude Fable 5 by 16.555.539+16.5
- LiveBench · ReasoningClaude Fable 5 by 15.289.774.5+15.2
- GPQA Diamondtie92.692.9tie
FactsEven
- AA-Omniscience · AccuracyClaude Fable 5 by 48.765.316.7+48.7
- AA-Omniscience · Non-hallucinationMiniMax M3 by 45.236.481.6+45.2
Images and chartsClaude Fable 5
- LMArena · VisionClaude Fable 5 by 71 rating points13251254+71 rating
- MMMU-ProClaude Fable 5 by 5.684.278.6+5.6
Following instructionsEven
- IFBenchMiniMax M3 by 19.463.582.9+19.4
- LiveBench · Instruction FollowingClaude Fable 5 by 18.375.857.5+18.3
Other results15 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.
- DeepSWEClaude Fable 5 by 507020+50
- AA-OmniscienceClaude Fable 5 by 41.943.31.4+41.9
- PaperBenchClaude Fable 5 by 36.288.852.6+36.2
- GDPval-AA v2Claude Fable 5 by 332 rating points17231391+332 rating
- AA IntelligenceClaude Fable 5 by 20.449.629.2+20.4
- AA Agentic IndexClaude Fable 5 by 20.25130.8+20.2
- Artificial Analysis Coding IndexClaude Fable 5 by 17.976.558.6+17.9
- livebench_languageClaude Fable 5 by 13.990.776.8+13.9
- τ³-Bench BankingClaude Fable 5 by 13.826.813+13.8
- OfficeQA ProClaude Fable 5 by 12.857.945.1+12.8
- τ²-Bench Telecom (AA run)Claude Fable 5 by 9.698.588.9+9.6
- LiveBenchClaude Fable 5 by 8.378.370+8.3
- Finance Agent v2Claude Fable 5 by 856.348.3+8
- livebench_data_analysisClaude Fable 5 by 4.380.576.2+4.3
- PostTrainBenchClaude Fable 5 by 4.341.437.1+4.3
Questions people ask
Which is better, Claude Fable 5 or MiniMax M3?
Claude Fable 5 wins five of the eight areas we test: coding, agents, reasoning, images and charts and math. MiniMax M3 wins none, but costs 98% less. They are level on facts, long documents and following instructions.
Which is better for coding?
Claude Fable 5. It wins 8 of the 8 coding tests both models report; MiniMax M3 wins none.
Which is cheaper?
Claude Fable 5 costs $10.00 per million input tokens and $50.00 per million output tokens; MiniMax M3 costs $0.30 and $1.20. That makes MiniMax M3 about 98% cheaper for the same work.
How do you compare the two?
We use the 43 benchmark tests both models have published scores on. The verdict counts the 28 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 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.