A clearer view of intelligence.
A free, independent, non-commercial guide to AI models and the technology behind them.
Technology moves.
Keep your perspective.
For people exploring AI.
For people building with it.
Artificials brings together an AI model directory, clear comparisons, editorial guides, a free AI course and free file tools. Our focus is making the information useful, understandable, and traceable to its source.
Artificials is a free, non-commercial project to help people understand AI. Everything on the site is free, with no paid plans, no advertising and no sponsored listings. No model maker pays to appear or to rank.
We are independent of the AI providers listed here. Provider names and logos identify their products and remain the property of their respective owners. A listing does not imply endorsement, partnership, or affiliation.
An open view of AI.
Explore models across providers with clear source and freshness information.
Evidence before a headline.
Read the methodology, distinguish missing data, and compare the metric that matters to you.
Free for everyone.
Rankings, guides, lessons and tools are free, without ads, accounts or paid plans.
Where the information comes from
Our catalog brings together public model specifications and pricing, refreshed automatically from each source. A refresh does not establish when every underlying fact was verified.
Coverage follows the source and can lag behind provider announcements. Missing prices, release dates, or capabilities remain unknown. We do not fill gaps with guessed values or treat an unreported result as zero. Confirm important details with the provider.
How to read the rankings
Capability rankings use published results from Epoch AI, LiveBench, T2I-CoReBench, SWE-bench, and DeepSWE and CursorBench as compiled by Epoch AI. Coding agent results come from Terminal-Bench 2.0 and SWE-bench. Each view identifies its dataset, evaluation scope and aggregation method. Missing evaluations are not scored zero. Benchmark results remain separate from specification and price rankings; a larger context window does not establish higher intelligence. See the leaderboard for the current metric and coverage.
Charts, highlights and the release timeline on the home page are calculated by Artificials from these sources. Each highlight names the test it comes from and says whether it is a test result, listing data or a calculation. Prices are list prices, not the cost of a particular task. Speed and latency are not shown, because no reliable public source is available for us to republish.