Compare Models
Select models (max 5)
DeepSeek V4MiniMax-M3
 |  |  |
|---|---|---|
| Vals IndexGDP-weighted benchmark | 38.63%±1.16 | 36.53%±1.17 |
| Legal Research BenchAgentic US legal research | 23.08%±2.93 | 29.81%±3.18 |
| Finance Agent (v2)Core financial analyst tasks | 44.08%±0.65 | 48.27%±0.44 |
| Tax Agent BenchAgentic US corporate tax research | 58.50%±3.13 | 49.69%±3.25 |
| MedCodeMedical billing code support | 40.45%±2.12 | 46.29%±2.10 |
| Code MigrationRewriting programs in new languages | 26.20%±4.04 | 19.93%±3.94 |
| Terminal-Bench 4.0Frontier-difficulty terminal tasks | 11.11%±1.34 | 1.01%±1.01 |
| Vibe Code Bench v1.1Building web apps from scratch | 49.93%±4.77 | 47.57%±5.44 |
Benchmarks
Vals Index *
DeepSeek V4
0.00%± 1.16
(43/43)MiniMax-M3
0.00%± 1.17
(43/43)Legal Research Bench *
DeepSeek V4
0.00%± 2.93
(73/73)MiniMax-M3
0.00%± 3.18
(73/73)Finance Agent (v2) *
DeepSeek V4
0.00%± 0.65
(74/74)MiniMax-M3
0.00%± 0.44
(74/74)Tax Agent Bench *
DeepSeek V4
0.00%± 3.13
(65/65)MiniMax-M3
0.00%± 3.25
(65/65)MedCode *
DeepSeek V4
0.00%± 2.12
(104/104)MiniMax-M3
0.00%± 2.10
(104/104)Code Migration *
DeepSeek V4
0.00%± 4.04
(73/73)MiniMax-M3
0.00%± 3.94
(73/73)Terminal-Bench 4.0
DeepSeek V4
0.00%± 1.34
(43/43)MiniMax-M3
0.00%± 1.01
(43/43)Vibe Code Bench v1.1 *
DeepSeek V4
0.00%± 4.77
(108/108)MiniMax-M3
0.00%± 5.44
(108/108)Overall performance
Performance on the Vals Index, a GDP-weighted aggregation of tasks across finance, coding, and law
•2026-10-02
Comparison by Industry
Model performance on different sections of the economy.
Legal
35.72%
vs.39.80%
Finance
56.57%
vs.57.36%
Healthcare
57.80%
vs.66.77%
Math
16.00%
vs.18.00%
Academic
88.32%
vs.86.02%
Education
N/A
vs.50.57%
Coding
43.34%
vs.40.91%
Cyber
N/A
vs.65.00%
Social Mobility
62.92%
vs.64.14%
Cost Analysis
Per task and per token model pricing.
Vals Index · USD
$1.306
$3.082
Cost / TestVals Index
$1.306
vs.$3.082
Input Cost/ 1M Tokens
$1.32
vs.$0.60
Input Cache Read/ 1M Tokens
$0.044
vs.$0.12
Output Cost/ 1M Tokens
$3.96
vs.$2.40
Model Metadata
Basic information about each model.
| | |
|---|---|---|
| Model provider | DeepSeek | MiniMax |
| Latency | 1928.91s | 2334.66s |
| Cost (In/Out) | $1.32 / $3.96 | $0.6 / $2.4 |
| Context Window | 1M | 1M |
| Max Output Token | 384k | 512k |
| Input Modality |