Compare Models

Select models (max 7)
DeepSeek V4.1 FlashHy4 Preview
Benchmarks

Vals Index *

DeepSeek V4.1 Flash
N/A
Hy4 Preview
0.00%± 1.13
(41/41)

Legal Research Bench *

DeepSeek V4.1 Flash
0.00%± 3.42
(98/98)
Hy4 Preview
0.00%± 3.46
(98/98)

Finance Agent v2.1 *

DeepSeek V4.1 Flash
0.00%± 2.02
(44/44)
Hy4 Preview
0.00%± 2.06
(44/44)

Tax Agent Bench *

DeepSeek V4.1 Flash
0.00%± 3.14
(89/89)
Hy4 Preview
0.00%± 3.38
(89/89)

MedCode *

DeepSeek V4.1 Flash
0.00%± 2.04
(107/107)
Hy4 Preview
0.00%± 2.13
(107/107)

Terminal-Bench-Science

DeepSeek V4.1 Flash
0.00%± 0.00
(47/47)
Hy4 Preview
0.00%± 1.43
(47/47)

Code Migration v1.1 *

DeepSeek V4.1 Flash
0.00%± 2.66
(30/30)
Hy4 Preview
0.00%± 3.59
(30/30)

Terminal-Bench 4.0

DeepSeek V4.1 Flash
0.00%± 1.75
(68/68)
Hy4 Preview
0.00%± 1.34
(68/68)

Vibe Code Bench v1.1 *

DeepSeek V4.1 Flash
0.00%± 2.89
(132/132)
Hy4 Preview
0.00%± 4.04
(132/132)

Overall performance

Performance on the Vals Index, a GDP-weighted aggregation of tasks across finance, coding, legal, and tax

2026-11-16

Comparison by Industry

Model performance on different sections of the economy.

Legal
63.42%
45.90%
Finance
53.27%
54.40%
Healthcare
63.34%
63.42%
Math
54.00%
75.00%
Science
29.65%
35.34%
Education
85.38%
N/A
Coding
45.77%
38.05%
Cyber
37.23%
33.82%
Social Mobility
69.99%
68.61%

Cost Analysis

Per task and per token model pricing.

Vals Index · USD
N/A
$1.568
Cost / TestVals Index
N/A
$1.568
Input Cost/ 1M Tokens
$0.30
$0.834
Input Cache Read/ 1M Tokens
$0.006
$0.042
Output Cost/ 1M Tokens
$1.20
$2.501

Model Metadata

Basic information about each model.

Model provider
DeepSeekDeepSeek
TencentOpenrouter
LatencyN/A3562.64s
Cost (In/Out)$0.3 / $1.2$0.834 / $2.501
Context Window1M1M
Max Output Token384k64k
Input Modality