Vals Index
Updated 11/18/2025
Vals Index
Benchmark consisting of a weighted performance across finance, law and coding tasks. Showing the potential impact that LLM's can have on the economy.
Top Model:
Claude Sonnet 4.5 (Thinking)
Number of models tested
24
Updated 11/18/2025
NewVals Multimodal Index
Benchmark consisting of a weighted performance across finance, law, coding, and education tasks. Showing the potential impact that LLM's can have on the economy.
Top Model:
Claude Sonnet 4.5 (Thinking)
Number of models tested
12
Legal Benchmarks
Updated 11/18/2025
CaseLaw (v2)
Private question-answer benchmark over Canadian court-cases.
Top Model:
GPT 5 Mini
Number of models tested
44
Updated 11/13/2025
LegalBench
Evaluating language models on a wide range of open source legal reasoning tasks.
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
93
Finance Benchmarks
Updated 11/17/2025
CorpFin (v2)
A private benchmark evaluating understanding of long-context credit agreements
Top Model:
Grok 4 Fast (Reasoning)
Number of models tested
62
Updated 11/18/2025
Finance Agent
Evaluating agents on core financial analyst tasks
Top Model:
GPT 5.1
Number of models tested
48
Updated 11/18/2025
MortgageTax
Evaluating reading and understanding tax certificates as images
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
49
Updated 11/13/2025
TaxEval (v2)
A Vals-created set of questions and responses to tax questions
Top Model:
Grok 3
Number of models tested
78
Healthcare Benchmarks
Updated 11/13/2025
MedQA
Evaluating language model bias in medical questions.
Top Model:
o1
Number of models tested
74
Math Benchmarks
Updated 11/13/2025
AIME
Challenging national math exam given to top high-school students
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
68
Updated 11/13/2025
MGSM
A multilingual benchmark for mathematical questions.
Top Model:
Claude Opus 4.1 (Thinking)
Number of models tested
70
Updated 08/26/2025
MATH 500
Academic math benchmark on probability, algebra, and trigonometry
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
54
Academic Benchmarks
Updated 11/18/2025
GPQA
Graduate-level Google-Proof Q&A benchmark evaluating models on questions that require deep reasoning.
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
70
Updated 11/13/2025
MMLU Pro
Academic multiple-choice benchmark covering 14 subjects including STEM, humanities, and social sciences.
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
68
Updated 11/13/2025
MMMU
Multimodal Multi-task Benchmark
Top Model:
GPT 5.1
Number of models tested
45
Education Benchmarks
Updated 11/18/2025
SAGE
Student Assessment with Generative Evaluation
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
26
Coding Benchmarks
Updated 11/13/2025
IOI
Based on the International Olympiad in Informatics
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
25
Updated 11/14/2025
LiveCodeBench
Our Implementation of the LiveCodeBench benchmark
Top Model:
GPT 5 Mini
Number of models tested
68
Updated 11/19/2025
SWE-bench
Solving production software engineering tasks
Top Model:
Gemini 3 Pro (11/25)
Number of models tested
31
Updated 11/19/2025
Terminal-Bench
State-of-the-art set of difficult terminal-based tasks
Top Model:
Claude Sonnet 4.5 (Thinking)
Number of models tested
35