Compare Models
Select models (max 7)
MiniMax-M3GPT 5.4 Mini
| Legal Research BenchAgentic US legal research | 29.81%±3.18 | 12.50%±2.30 |
| Finance Agent v2.1Core financial analyst tasks | 54.34%±2.02 | N/A |
| Tax Agent BenchAgentic US corporate tax research | 20.59%±2.87 | 8.81%±2.04 |
| MedCodeMedical billing code support | 46.29%±2.10 | N/A |
| Code Migration v1.1Rewriting programs in new languages | 20.90%±3.44 | N/A |
| Terminal-Bench 4.0Frontier-difficulty terminal tasks | 1.01%±1.01 | 2.52%±1.01 |
| Vibe Code Bench v1.1Building web apps from scratch | 47.57%±5.44 | 47.97%±5.61 |
Benchmarks
Legal Research Bench *
MiniMax-M3
0.00%± 3.18
(98/98)GPT 5.4 Mini
0.00%± 2.30
(98/98)Finance Agent v2.1 *
MiniMax-M3
0.00%± 2.02
(44/44)GPT 5.4 Mini
N/ATax Agent Bench *
MiniMax-M3
0.00%± 2.87
(89/89)GPT 5.4 Mini
0.00%± 2.04
(89/89)MedCode *
MiniMax-M3
0.00%± 2.10
(107/107)GPT 5.4 Mini
N/ACode Migration v1.1 *
MiniMax-M3
0.00%± 3.44
(30/30)GPT 5.4 Mini
N/ATerminal-Bench 4.0
MiniMax-M3
0.00%± 1.01
(68/68)GPT 5.4 Mini
0.00%± 1.01
(68/68)Vibe Code Bench v1.1 *
MiniMax-M3
0.00%± 5.44
(132/132)GPT 5.4 Mini
0.00%± 5.61
(132/132)Overall performance
Performance on the Vals Index, a GDP-weighted aggregation of tasks across finance, coding, legal, and tax