o3 is the stronger all-rounder.
Scores updated · 24 tests both models report · How we compare
Where each one wins
Tests won in each of the four areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
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.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where o3 pulls ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+34.8points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+30.2points ahead
- Reasons across sets of long documentsAA-LCR+19points ahead
Where MiniMax M1 40K pulls ahead
- Questions about very long textsLongBench v2+2.2points ahead
Every test, side by side
All 24 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Codingo3
- Terminal-Bench Hardo3 by 34.82.337.1+34.8
- SWE-bench Verifiedo3 by 13.555.669.1+13.5
- SciCodeo3 by 3.237.841+3.2
Reasoningo3
- GPQA Diamondo3 by 14.568.282.7+14.5
- Humanity's Last Examo3 by 12.37.820.1+12.3
Long documentsEven
- AA-LCRo3 by 1955.774.7+19
- LongBench v2MiniMax M1 40K by 2.26158.8+2.2
Following instructionso3
- IFBencho3 by 30.241.271.4+30.2
- Multi-Challengeo3 by 15.744.760.4+15.7
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.
- τ²-Bench Telecom (AA run)o3 by 49.131.680.7+49.1
- SimpleQAo3 by 31.517.949.4+31.5
- Artificial Analysis Coding Indexo3 by 24.314.138.4+24.3
- LiveCodeBencho3 by 22.462.384.7+22.4
- OpenAI-MRCR (128k)MiniMax M1 40K by 19.676.156.5+19.6
- ZebraLogico3 by 15.780.195.8+15.7
- LiveCodeBench (24/8~25/5)o3 by 13.562.375.8+13.5
- AIME 2025o3 by 11.874.686.4+11.8
- AA Intelligenceo3 by 10.21020.2+10.2
- AIME 2024o3 by 8.383.391.6+8.3
- TAU-bench (airline)MiniMax M1 40K by 86052+8
- TAU-bench (retail)o3 by 6.167.873.9+6.1
- MMLU-Proo3 by 4.480.685+4.4
- MATH-500 (EM)o3 by 2.19698.1+2.1
- FullStackBencho3 by 1.767.669.3+1.7
Questions people ask
Which is better, MiniMax M1 40K or o3?
o3 wins three of the four areas where both have results: coding, reasoning and following instructions. MiniMax M1 40K wins none. They are level on long documents.
Which is better for coding?
o3. It wins 3 of the 3 coding tests both models report; MiniMax M1 40K wins none.
How do you compare the two?
We use the 24 benchmark tests both models have published scores on. The verdict counts the 9 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.