Compare Models
Select models (max 5)
DeepSeek V4 Flash 0731GLM 5.3
 |  |  |
|---|---|---|
| Vals IndexGDP-weighted benchmark | 47.96%±1.24 | 53.51%±1.30 |
| Legal Research BenchAgentic US legal research | 30.29%±3.19 | 49.04%±3.48 |
| Finance Agent (v2)Core financial analyst tasks | 49.52%±0.47 | 55.84%±2.07 |
| Tax Agent BenchAgentic US corporate tax research | 62.56%±3.14 | 73.09%±2.97 |
| MedCodeMedical billing code support | 41.41%±2.15 | 42.86%±2.11 |
| Terminal-Bench ScienceExpert-authored scientific research workflows | 4.29%±2.44 | 5.71%±2.79 |
| Code MigrationRewriting programs in new languages | 38.63%±4.38 | 44.22%±4.29 |
| Terminal-Bench 4.0Frontier-difficulty terminal tasks | 18.69%±3.54 | 38.89%±1.82 |
| Vibe Code Bench v1.1Building web apps from scratch | 74.74%±3.44 | 78.13%±3.94 |
Benchmarks
Vals Index *
DeepSeek V4 Flash 0731
0.00%± 1.24
(43/43)GLM 5.3
0.00%± 1.30
(43/43)Legal Research Bench *
DeepSeek V4 Flash 0731
0.00%± 3.19
(73/73)GLM 5.3
0.00%± 3.48
(73/73)Finance Agent (v2) *
DeepSeek V4 Flash 0731
0.00%± 0.47
(74/74)GLM 5.3
0.00%± 2.07
(74/74)Tax Agent Bench *
DeepSeek V4 Flash 0731
0.00%± 3.14
(65/65)GLM 5.3
0.00%± 2.97
(65/65)MedCode *
DeepSeek V4 Flash 0731
0.00%± 2.15
(104/104)GLM 5.3
0.00%± 2.11
(104/104)Terminal-Bench Science
DeepSeek V4 Flash 0731
0.00%± 2.44
(38/38)GLM 5.3
0.00%± 2.79
(38/38)Code Migration *
DeepSeek V4 Flash 0731
0.00%± 4.38
(73/73)GLM 5.3
0.00%± 4.29
(73/73)Terminal-Bench 4.0
DeepSeek V4 Flash 0731
0.00%± 3.54
(43/43)GLM 5.3
0.00%± 1.82
(43/43)Vibe Code Bench v1.1 *
DeepSeek V4 Flash 0731
0.00%± 3.44
(108/108)GLM 5.3
0.00%± 3.94
(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
38.78%
vs.47.40%
Finance
59.94%
vs.64.41%
Healthcare
60.89%
vs.65.84%
Math
56.00%
vs.49.00%
Science
34.37%
vs.14.34%
Academic
88.05%
vs.87.45%
Coding
48.94%
vs.52.74%
Cyber
72.08%
vs.72.08%
Social Mobility
N/A
vs.68.54%
Cost Analysis
Per task and per token model pricing.
Vals Index · USD
$0.848
$7.249
Cost / TestVals Index
$0.848
vs.$7.249
Input Cost/ 1M Tokens
$0.44
vs.$1.40
Input Cache Read/ 1M Tokens
$0.014
vs.$0.26
Output Cost/ 1M Tokens
$1.32
vs.$4.40
Model Metadata
Basic information about each model.
| | |
|---|---|---|
| Model provider | DeepSeek | zAI |
| Latency | 2585.58s | 4392.77s |
| Cost (In/Out) | $0.44 / $1.32 | $1.4 / $4.4 |
| Context Window | 1M | 1M |
| Max Output Token | 384k | 128k |
| Input Modality |