Compare Models
Select models (max 7)
Claude Haiku 5.5Gemini 3.8 Flash
 |  |  |
|---|---|---|
| Vals IndexGDP-weighted benchmark | 58.71%±1.28 | 55.95%±1.14 |
| Legal Research BenchAgentic US legal research | 43.27%±3.44 | 38.94%±3.39 |
| Finance Agent v2.1Core financial analyst tasks | 65.37%±2.06 | 65.76%±2.04 |
| Tax Agent BenchAgentic US corporate tax research | 31.92%±3.35 | 32.12%±3.37 |
| MedCodeMedical billing code support | 51.54%±2.17 | 48.13%±2.18 |
| Terminal-Bench-ScienceExpert-authored scientific research workflows | 10.00%±3.61 | 8.57%±3.37 |
| Code Migration v1.1Rewriting programs in new languages | 52.36%±3.49 | 38.65%±3.65 |
| Terminal-Bench 4.0Frontier-difficulty terminal tasks | 36.36%±2.31 | 19.70%±2.31 |
| Vibe Code Bench v1.1Building web apps from scratch | 90.44%±1.73 | 78.65%±3.88 |
Benchmarks
Vals Index *
Claude Haiku 5.5
0.00%± 1.28
(26/26)Gemini 3.8 Flash
0.00%± 1.14
(26/26)Legal Research Bench *
Claude Haiku 5.5
0.00%± 3.44
(84/84)Gemini 3.8 Flash
0.00%± 3.39
(84/84)Finance Agent v2.1 *
Claude Haiku 5.5
0.00%± 2.06
(30/30)Gemini 3.8 Flash
0.00%± 2.04
(30/30)Tax Agent Bench *
Claude Haiku 5.5
0.00%± 3.35
(75/75)Gemini 3.8 Flash
0.00%± 3.37
(75/75)MedCode *
Claude Haiku 5.5
0.00%± 2.17
(107/107)Gemini 3.8 Flash
0.00%± 2.18
(107/107)Terminal-Bench-Science
Claude Haiku 5.5
0.00%± 3.61
(42/42)Gemini 3.8 Flash
0.00%± 3.37
(42/42)Code Migration v1.1 *
Claude Haiku 5.5
0.00%± 3.49
(28/28)Gemini 3.8 Flash
0.00%± 3.65
(28/28)Terminal-Bench 4.0
Claude Haiku 5.5
0.00%± 2.31
(54/54)Gemini 3.8 Flash
0.00%± 2.31
(54/54)Vibe Code Bench v1.1 *
Claude Haiku 5.5
0.00%± 1.73
(118/118)Gemini 3.8 Flash
0.00%± 3.88
(118/118)Overall performance
Performance on the Vals Index, a GDP-weighted aggregation of tasks across finance, coding, and law
•2026-10-22
Comparison by Industry
Model performance on different sections of the economy.
Legal
23.72%
vs.46.01%
Finance
65.69%
vs.62.11%
Healthcare
70.54%
vs.66.32%
Math
89.00%
vs.48.00%
Science
38.50%
vs.35.76%
Academic
N/A
vs.91.25%
Education
45.78%
vs.35.06%
Coding
41.90%
vs.48.20%
Cyber
40.60%
vs.27.41%
Social Mobility
63.19%
vs.65.29%
Cost Analysis
Per task and per token model pricing.
Vals Index · USD
$3.173
$6.065
Cost / TestVals Index
$3.173
vs.$6.065
Input Cost/ 1M Tokens
$0.10
vs.$1.50
Input Cache Write/ 1M Tokens
$0.125
vs.N/A
Input Cache Read/ 1M Tokens
$0.01
vs.$0.15
Output Cost/ 1M Tokens
$0.50
vs.$7.50
Model Metadata
Basic information about each model.
| | |
|---|---|---|
| Model provider | Anthropic | Google |
| Latency | 2988.33s | 2158.48s |
| Cost (In/Out) | $0.1 / $0.5 | $1.5 / $7.5 |
| Context Window | 1M | 1M |
| Max Output Token | 128k | 66k |
| Input Modality |