Every model, one reference.
Releases by day · 713 in the last 12 months · drag to scrollswipe to scroll
OpenAIAnthropicGooglexAIOther labs
Days since last release
OpenAI8GPT-Live-1.1
Anthropic16Claude Haiku 5.5
Google17EmbeddingGemma 2
xAI9Grok Voice 3.0
Best overall today#4 of 146 Q/$
Cheapest frontier$1.20 / 1M out
Freshest update2026-10-23
Editors' picks · featured
See all 18 →Where most teams start
DevelopersCoding
EC80.3% SWE-bench Pro, 96% SWE-bench Verified on Vals.ai, and 85.0% OSWorld-Verified make it the best production coding pick for non-trivial engineering tasks.
Claude Fable 5.1
Claude Opus 5.5
7 picks · 85 eligibleView all →
DevelopersAgents
ECEveryday agent at $2 / $10 per 1M tokens, with a 1M window, 81.2% OSWorld-Verified and 86.6% BrowseComp multi-agent.
Claude Opus 5.5
Claude Opus 5
7 picks · 161 eligibleView all →
Knowledge workersWriting
ECWrites publishable paragraphs, follows tone notes closely and edits well. $5 / $25 per 1M tokens.
Claude Opus 5.5
GPT-5.5
5 picks · 9 eligibleView all →
Knowledge workersResearch
ECAnthropic reports finance, trading and analytics results for Fable 5, with a 1M window at $10 / $50 per 1M tokens.
Claude Fable 5.1
GPT-6 Astra
7 picks · 13 eligibleView all →
CreativesImage
ECFLUX.2 Klein 9B is a 9B distilled text-to-image model on fal. Not FLUX.2 Dev.
GPT Image 2
Midjourney V6
4 picks · 10 eligibleView all →
CreativesVideo
ECBest overall video quality in the catalog: 4-, 6- or 8-second clips with native audio, on Vertex AI, Vercel AI Gateway and OpenRouter.
Runway Gen-4.5
Wan 2.7
4 picks · 30 eligibleView all →
All 18 boards
Every leaderboard, by audience
Cheat sheet