Compare Models

Select models (max 7)
Inkling 2 SmallHy4
Benchmarks

Vals Index *

Inkling 2 Small
0.00%± 1.18
(34/34)
Hy4
0.00%± 1.08
(34/34)

Legal Research Bench *

Inkling 2 Small
0.00%± 3.47
(91/91)
Hy4
0.00%± 3.42
(91/91)

Finance Agent v2.1 *

Inkling 2 Small
0.00%± 2.02
(37/37)
Hy4
0.00%± 2.05
(37/37)

Tax Agent Bench *

Inkling 2 Small
0.00%± 3.27
(82/82)
Hy4
0.00%± 3.30
(82/82)

Terminal-Bench-Science

Inkling 2 Small
0.00%± 2.01
(44/44)
Hy4
N/A

Terminal-Bench 4.0

Inkling 2 Small
0.00%± 3.07
(61/61)
Hy4
0.00%± 0.88
(61/61)

Vibe Code Bench v1.1 *

Inkling 2 Small
0.00%± 3.45
(125/125)
Hy4
0.00%± 3.55
(125/125)

Overall performance

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

2026-11-05

Comparison by Industry

Model performance on different sections of the economy.

Legal
34.15%
26.92%
Finance
58.65%
55.09%
Science
36.43%
N/A
Education
88.57%
N/A
Coding
34.45%
60.20%
Cyber
37.32%
N/A

Cost Analysis

Per task and per token model pricing.

Vals Index · USD
$3.509
$1.72
Cost / TestVals Index
$3.509
$1.72
Input Cost/ 1M Tokens
$0.30
$0.834
Input Cache Read/ 1M Tokens
$0.03
$0.042
Output Cost/ 1M Tokens
$1.00
$2.501

Model Metadata

Basic information about each model.

Model provider
Thinking MachinesThinking Machines
TencentOpenrouter
Latency3034.53s3091.47s
Cost (In/Out)$0.3 / $1$0.834 / $2.501
Context Window1M1M
Max Output Token1M66k
Input Modality