Compare Models
Select models (max 5)
DeepSeek V4.1 FlashGPT-6 Luna
 |  |  |
|---|---|---|
| Vals IndexGDP-weighted benchmark | 51.32%±1.13 | 51.22%±1.07 |
| Legal Research BenchAgentic US legal research | 41.35%±3.42 | 30.29%±3.19 |
| Finance Agent (v2)Core financial analyst tasks | 53.48%±0.39 | 49.87%±0.23 |
| Tax Agent BenchAgentic US corporate tax research | 62.46%±3.15 | 58.86%±3.21 |
| MedCodeMedical billing code support | 41.17%±2.04 | 44.69%±2.30 |
| Terminal-Bench ScienceExpert-authored scientific research workflows | 0.00%±0.00 | 4.29%±2.44 |
| Code MigrationRewriting programs in new languages | 45.62%±4.29 | 42.55%±4.42 |
| Terminal-Bench 4.0Frontier-difficulty terminal tasks | 19.70%±1.75 | 13.64%±1.51 |
| Vibe Code Bench v1.1Building web apps from scratch | 84.74%±2.89 | 81.65%±3.38 |
Benchmarks
Vals Index *
DeepSeek V4.1 Flash
0.00%± 1.13
(43/43)GPT-6 Luna
0.00%± 1.07
(43/43)Legal Research Bench *
DeepSeek V4.1 Flash
0.00%± 3.42
(73/73)GPT-6 Luna
0.00%± 3.19
(73/73)Finance Agent (v2) *
DeepSeek V4.1 Flash
0.00%± 0.39
(74/74)GPT-6 Luna
0.00%± 0.23
(74/74)Tax Agent Bench *
DeepSeek V4.1 Flash
0.00%± 3.15
(65/65)GPT-6 Luna
0.00%± 3.21
(65/65)MedCode *
DeepSeek V4.1 Flash
0.00%± 2.04
(104/104)GPT-6 Luna
0.00%± 2.30
(104/104)Terminal-Bench Science
DeepSeek V4.1 Flash
0.00%± 0.00
(38/38)GPT-6 Luna
0.00%± 2.44
(38/38)Code Migration *
DeepSeek V4.1 Flash
0.00%± 4.29
(73/73)GPT-6 Luna
0.00%± 4.42
(73/73)Terminal-Bench 4.0
DeepSeek V4.1 Flash
0.00%± 1.75
(43/43)GPT-6 Luna
0.00%± 1.51
(43/43)Vibe Code Bench v1.1 *
DeepSeek V4.1 Flash
0.00%± 2.89
(108/108)GPT-6 Luna
0.00%± 3.38
(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
43.76%
vs.16.60%
Finance
57.72%
vs.59.08%
Healthcare
63.34%
vs.64.20%
Math
54.00%
vs.64.00%
Science
29.65%
vs.28.38%
Education
47.88%
vs.48.09%
Coding
39.57%
vs.38.78%
Cyber
37.23%
vs.76.25%
Social Mobility
64.28%
vs.57.65%
Cost Analysis
Per task and per token model pricing.
Vals Index · USD
$0.332
$0.431
Cost / TestVals Index
$0.332
vs.$0.431
Input Cost/ 1M Tokens
$0.30
vs.$0.10
Input Cache Write/ 1M Tokens
N/A
vs.$0.125
Input Cache Read/ 1M Tokens
$0.006
vs.$0.01
Output Cost/ 1M Tokens
$1.20
vs.$0.50
Model Metadata
Basic information about each model.
| | |
|---|---|---|
| Model provider | DeepSeek | OpenAI |
| Latency | 1673.17s | 1833.68s |
| Cost (In/Out) | $0.3 / $1.2 | $0.1 / $0.5 |
| Context Window | 1M | 1M |
| Max Output Token | 384k | 128k |
| Input Modality |