Claude Opus 4.8 vs MiniMax M3
Wins 5 of 8 areas
Coding · Agents · Reasoning · Images and charts · Math
Wins 1 of 8 areas
Long documents
Claude Opus 4.8 is the stronger all-rounder.MiniMax M3 is cheaper and better at long documents.
Scores updated · 45 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 Opus 4.88 of 8 tests
- AgentsCarrying out multi-step tasks on its own60Claude Opus 4.86 of 7 tests · 1 tie
- ReasoningHard problems that need careful thinking40Claude Opus 4.84 of 5 tests · 1 tie
- MathCompetition and research-level math20Claude Opus 4.82 of 2 tests
- Images and chartsUnderstanding pictures, charts and video10Claude Opus 4.81 of 1 test
- Long documentsFinding answers in very long texts01MiniMax M31 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
- Following instructionsDoing exactly what it is asked11Even1 each
Images and charts 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 95% 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 4.8 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+32.1points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+19.7points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+19.4points ahead
Where MiniMax M3 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+20.9points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+20.7points ahead
- Reasons across sets of long documentsAA-LCR+5.3points ahead
Every test, side by side
All 45 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 4.8
- Terminal-Bench 2.1Claude Opus 4.8 by 19.484.665.2+19.4
- Terminal-Bench HardClaude Opus 4.8 by 15.958.342.4+15.9
- LiveBench · CodingClaude Opus 4.8 by 13.681.868.2+13.6
- SWE-bench ProClaude Opus 4.8 by 10.269.259+10.2
- LiveBench · Agentic CodingClaude Opus 4.8 by 9.850.540.7+9.8
- SWE-bench VerifiedClaude Opus 4.8 by 8.188.680.5+8.1
- LMArena · WebDevClaude Opus 4.8 by 74 rating points15561482+74 rating
- SciCodeClaude Opus 4.8 by 7.354.447.1+7.3
AgentsClaude Opus 4.8
- Terminal-Bench 4.0Claude Opus 4.8 by 19.721.72+19.7
- τ-Bench V3 · BankingClaude Opus 4.8 by 18.934.215.3+18.9
- OSWorld-VerifiedClaude Opus 4.8 by 13.383.470.1+13.3
- AA ApexAgentsClaude Opus 4.8 by 11.739.427.7+11.7
- GDPValClaude Opus 4.8 by 10.547.837.3+10.5
- MCP AtlasClaude Opus 4.8 by 882.274.2+8
- BrowseComptie84.383.5tie
ReasoningClaude Opus 4.8
- SimpleBenchClaude Opus 4.8 by 1964.845.8+19
- CritPtClaude Opus 4.8 by 17.220.93.7+17.2
- LiveBench · ReasoningClaude Opus 4.8 by 14.789.274.5+14.7
- Humanity's Last ExamClaude Opus 4.8 by 9.748.739+9.7
- GPQA Diamondtie9292.9tie
FactsEven
- AA-Omniscience · AccuracyClaude Opus 4.8 by 32.148.816.7+32.1
- AA-Omniscience · Non-hallucinationMiniMax M3 by 20.960.781.6+20.9
Images and chartsClaude Opus 4.8
- LMArena · VisionClaude Opus 4.8 by 41 rating points12951254+41 rating
MathClaude Opus 4.8
- LiveBench · MathematicsClaude Opus 4.8 by 17.394.377+17.3
- HMMT Feb. 2026Claude Opus 4.8 by 11.195.584.4+11.1
Following instructionsEven
- IFBenchMiniMax M3 by 20.762.282.9+20.7
- LiveBench · Instruction FollowingClaude Opus 4.8 by 14.57257.5+14.5
Other results17 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 Opus 4.8 by 385820+38
- PaperBenchClaude Opus 4.8 by 27.780.352.6+27.7
- NL2RepoClaude Opus 4.8 by 27.669.742.1+27.6
- AA-OmniscienceClaude Opus 4.8 by 27.428.81.4+27.4
- OfficeQA ProClaude Opus 4.8 by 21.166.245.1+21.1
- GDPval-AA v2Claude Opus 4.8 by 202 rating points15931391+202 rating
- Artificial Analysis Coding IndexClaude Opus 4.8 by 15.774.358.6+15.7
- AA IntelligenceClaude Opus 4.8 by 12.641.829.2+12.6
- HMMT Nov. 2025Claude Opus 4.8 by 12.196.584.4+12.1
- AA Agentic IndexClaude Opus 4.8 by 11.842.630.8+11.8
- livebench_data_analysisMiniMax M3 by 10.26676.2+10.2
- LiveBenchClaude Opus 4.8 by 7.277.270+7.2
- Tool-DecathlonClaude Opus 4.8 by 7.159.952.8+7.1
- Finance Agent v2Claude Opus 4.8 by 5.653.948.3+5.6
- τ²-Bench Telecom (AA run)Claude Opus 4.8 by 5.594.488.9+5.5
- livebench_languageClaude Opus 4.8 by 2.979.776.8+2.9
- PostTrainBenchtie37.237.1tie
Questions people ask
Which is better, Claude Opus 4.8 or MiniMax M3?
Claude Opus 4.8 wins five of the eight areas we test: coding, agents, reasoning, images and charts and math. MiniMax M3 wins long documents, and costs 95% less. They are level on facts and following instructions.
Which is better for coding?
Claude Opus 4.8. It wins 8 of the 8 coding tests both models report; MiniMax M3 wins none.
Which is cheaper?
Claude Opus 4.8 costs $5.00 per million input tokens and $25.00 per million output tokens; MiniMax M3 costs $0.30 and $1.20. That makes MiniMax M3 about 95% cheaper for the same work.
How do you compare the two?
We use the 45 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 17 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.