LM Council compares language models across benchmarks including math, science, and software engineering, with source links and checked dates.
Epoch AI
Epoch AI’s Capabilities Index combines scores from multiple benchmarks into a scale for comparing AI models over time.
About Epoch AI
Epoch AI's Epoch Capabilities Index (ECI) combines scores from many different AI benchmarks into a single general capability scale. This allows comparisons between models even over timespans long enough for individual benchmarks to reach saturation. The site presents the index through views such as score versus time, score versus training compute, a domain-versus-general view, and a leaderboard table listing models, labs and ECI values.
Key features
- Combines many AI benchmark scores into one general capability scale
- Charts score vs time and score vs training compute
- Leaderboard table of models, labs and ECI values
- Filter by benchmark, domain, saturation level and creator
- Color by country, organization or accessibility; frontier-trend-only option
- Interpreting chart converts an ECI value into expected benchmark scores
Best for: Researchers, analysts and writers who need to compare AI model capability across many benchmarks or across long time periods.
Good for
- Covers comparisons over spans where single benchmarks saturate
- Table can be downloaded and is marked CC-BY
- Filtering by domain and saturation level supports focused views
Watch out for
- An ECI value on its own is an abstract number needing interpretation
- Coverage is limited to the benchmarks selected into the index
- No pricing information is provided
Epoch AI at a glance
- Website
- epoch.ai/benchmarks/eci
- Type
- Online tool
- Pricing
- Not confirmed. Check the site
- Platforms
- Web
- Open source
- Not stated
- Language
- English
- Last checked
- Oct 5, 2026 · Online when checked
- Listed
- Oct 5, 2026
Epoch AI alternatives
Tools and sites you could use instead, ranked by how closely they match.
LLM Explorer indexes AI agents, frameworks, SDKs, and open-source language models.
The Wolfram LLM Benchmarking Project tracks model performance on a code-generation task written in Wolfram Language.
Promptomania combines a model directory, visual prompt builder, and guides for image, video, 3D, and audio generation.
CivArchive helps users search, verify, and track AI models across multiple platforms using SHA256 hashes.
IUMB benchmarks language models on undergraduate-level mathematics problems.
Popular in AI Directories & Resources
Well-known listings from the same topic.
getdesign.mdgetdesign.mdfuturepediafuturepedia.ioAwesomeclaudeawesomeclaude.aiAI Depotaidepot.coAI Valleyaivalley.aiAI Weeklyaiweekly.co
Epoch AI FAQ
What does an ECI value mean on its own?
An ECI value by itself is an abstract number. A separate chart converts it back into concrete benchmark performance, showing the expected score a model with a given ECI would receive, based on logistic fits to real (ECI, score) observations.
Can the underlying data be downloaded?
Yes. The table has a download option and is marked CC-BY.
Which benchmarks are included in the index?
The index draws on benchmarks from benchmark creators, model developers and Epoch AI, spanning domains such as agent tasks, computer use, conceptual reasoning, games, language understanding, long context, mathematics, multimodal tasks, science, software engineering, world knowledge, and writing and creativity.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.