AI
Every model.
Every release.
The frontier AI models today — Claude Opus 4.8, GPT-5.5, Gemini 3.5 Flash, Grok 4.3, Muse Spark, DeepSeek-V4-Pro, Mistral Medium 3.5, and Qwen3.7-Max — alongside the full release history of every lab from 2018 onward. The cards below open the cross-family rollups; per-family release timelines and surrounding company and legal history are linked from each.
Models
Every frontier-AI model family at a glance — Anthropic Claude, OpenAI ChatGPT, Google Gemini, xAI Grok, Meta Llama, DeepSeek, Mistral, Alibaba Qwen. Lab, current model, ship date, total versions. Sortable, filterable by lab status (US public / US private / Chinese / EU). Click into a family for the full lineage.
Open →Release cadence
How often each AI provider ships. Cross-family roll-up of the eight Versions pages on this site — rolling 12-month release count per provider, lifetime and last-12-month median interval, per-provider timeline, and a rolling-12-month trend chart over time. One headline number; the rest is the data behind it.
Open →Context windows
Every model’s context window in one table — input and output token limits, ship date, status. Plus a log-scale historical chart from GPT-3’s 2,048 tokens in 2020 to the 10M-token frontier today. Sortable, filterable by provider or minimum context size.
Open →Model lineage
Visual family trees of every major AI model line — how Claude 1 became Claude 4.7, how the GPT line and o-series merged into GPT-5, what’s a fine-tune vs. a new base. Each typed edge cites the provider model card, paper, or release post that documents the relationship.
Open →Benchmarks
Every frontier model on every major benchmark — ARC-AGI-2, GPQA Diamond, Humanity’s Last Exam, SWE-Bench Pro, AIME, MMMU. Each cell carries the lab-claimed score and the independent re-run side-by-side, with a flag when the two disagree by more than five points. Cited per cell.
Open →Pricing history
Per-token API pricing across every major frontier model, current and historical. Sortable comparison table, log-scale chart of the frontier price floor from GPT-3’s $60/M davinci tokens in 2020 to today’s sub-dollar floor, per-provider price-cut log. Cached and batch tiers documented alongside the headline rates.
Open →