Benchmarks

Standardized LLM evaluations across knowledge, coding, reasoning, and multimodal tasks.

Select Benchmark

MMLU

Massive Multitask Language Understanding - 57 subjects
Knowledge
1
claude-opus-4-6Anthropic
92.0%
2
o3OpenAI
91.8%
3
claude-sonnet-4-6Anthropic
90.5%
4
gemini-2.5-proGoogle
89.1%
5
deepseek-v3DeepSeek
88.5%
6
gpt-4oOpenAI
87.2%