Compare Models
Select models (max 7)
GPT 5 CodexMistral Large 4 OSS
 |  |  |
|---|---|---|
| Vals IndexGDP-weighted benchmark | N/A | 56.42%±1.26 |
| Legal Research BenchAgentic US legal research | N/A | 35.58%±3.33 |
| Finance Agent v2.1Core financial analyst tasks | N/A | 61.72%±2.05 |
| Tax Agent BenchAgentic US corporate tax research | N/A | 27.33%±2.95 |
| Terminal-Bench 4.0Frontier-difficulty terminal tasks | N/A | 26.26%±5.13 |
| Vibe Code Bench v1.1Building web apps from scratch | N/A | 81.56%±3.29 |
Benchmarks
Vals Index *
GPT 5 Codex
N/AMistral Large 4 OSS
0.00%± 1.26
(30/30)Legal Research Bench *
GPT 5 Codex
N/AMistral Large 4 OSS
0.00%± 3.33
(88/88)Finance Agent v2.1 *
GPT 5 Codex
N/AMistral Large 4 OSS
0.00%± 2.05
(34/34)Tax Agent Bench *
GPT 5 Codex
N/AMistral Large 4 OSS
0.00%± 2.95
(79/79)Terminal-Bench 4.0
GPT 5 Codex
N/AMistral Large 4 OSS
0.00%± 5.13
(58/58)Vibe Code Bench v1.1 *
GPT 5 Codex
N/AMistral Large 4 OSS
0.00%± 3.29
(122/122)Overall performance
Performance on the Vals Index, a GDP-weighted aggregation of tasks across finance, coding, and law
•2026-10-29
Comparison by Industry
Model performance on different sections of the economy.
Legal
N/A
vs.25.70%
Finance
N/A
vs.52.58%
Coding
84.72%
vs.53.91%
Cost Analysis
Per task and per token model pricing.
Vals Index · USD
N/A
$3.717
Cost / TestVals Index
N/A
vs.$3.717
Input Cost/ 1M Tokens
$1.25
vs.$0.70
Input Cache Read/ 1M Tokens
$0.125
vs.$0.07
Output Cost/ 1M Tokens
$10.00
vs.$2.10
Model Metadata
Basic information about each model.
| | |
|---|---|---|
| Model provider | OpenAI | Mistral AI |
| Latency | N/A | 2868.04s |
| Cost (In/Out) | $1.25 / $10 | $0.7 / $2.1 |
| Context Window | 400k | 1M |
| Max Output Token | 128k | 256k |
| Input Modality |