Benchmarks for frontier, agentic, and safety capabilities
Including benchmarks on agentic coding, frontier reasoning, and safety alignment.
From leading AI labs including OpenAI, Anthropic, Google, Meta, and open-source contributors.
HSS, in partnership with Elorian, is a benchmark for intuitive visual reasoning across diverse image and video inputs.
Gemini 4 Carbon (high)
NEW71.10±3.60
Claude Fable 5.2 (max)
NEW59.50±3.80
Gemini-4-Argon (high)
NEW55.40±4.20
We introduce DistressBench, designed to measure the quality of crisis support a model delivers.
muse-spark-1.2
88.39±0.93
muse-spark-open
NEW87.74±1.03
muse-wonder-1.0
NEW85.89±1.05
Challenging LLMs at the frontier of human knowledge
claude-fable-5-2
NEW72.40±2.80
60.60±3.00
55.00±3.10
DrugDiscoveryBench: 82 expert-curated tasks evaluating how reliably frontier coding agents perform the computational work of early-stage drug discovery.
Gemini 4 Carbon (mini-swe-agent) high
NEW71.50±2.10
Gemini 4 Hydrogen (mini-swe-agent) high
NEW69.10±3.50
GPT 6 Astra (mini-SWE-agent) max
NEW68.70±2.50
Evaluates whether agents recognize information gaps and ask targeted clarifying questions.
Claude Fable 5.2 (*public set only)
NEW80.00±4.95
Muse Wonder 1.0
79.67±4.17
GPT 6.1 Terra
68.00±5.02
Evaluating real-world tool use through the Model Context Protocol (MCP)
Gemini-4-Argon
NEW90.70±1.80
Muse Spark 1.1
88.10±1.95
Fable 5.1
87.20±2.05
Evaluating long-horizon software engineering tasks in public open source repositories
SWE Atlas evaluates coding agents on codebase comprehension, test writing and refactoring
Evaluating Professional Reasoning in Finance and Legal Practice
Evaluating AI agents ability to perform real-world, economically valuable remote work
Fable 5.2
NEW41.25
Opus 5.5
21.25
GPT 6 Astra
20.83
Frontier Risk Evaluation for National Security and Public Safety
5.78±1.26
7.64±1.15
8.24±1.93
We conduct high-complexity evaluations to expose model failures, prevent benchmark saturation, and push model capabilities--while continuously evaluating the latest frontier models.
Humans design complex evaluations and define precise criteria to assess models, while LLMs scale evaluations--ensuring efficiency and alignment with human judgment.
Our leaderboards are built on carefully curated evaluation sets, combining private datasets to prevent overfitting and open-source datasets for broad benchmarking and comparability.
If you'd like to add your model to this leaderboard or a future version, please contact [email protected]. To ensure leaderboard integrity, we require that models can only be featured the FIRST TIME when an organization encounters the prompts.