AI models we track
Every model that has entered the tracker, from first release to retirement, with the pricing and lifecycle details vendors bury in a launch PDF.
The homepage table only shows current models that already have at least one sourced benchmark score, so its rankings stay meaningful. This page is the full catalog — including models with no scores sourced yet and models that have since been superseded — so you can see everything that exists, not just what is currently rankable.
License terms and knowledge-cutoff dates live on each model’s own page — they vary too much in length and completeness to fit cleanly into a table row.
| Context | |||||
|---|---|---|---|---|---|
| GLM-5.3 Zhipu AI (Z.ai) | 2026-08-14 | $1.40 / $4.40 | 1M | 5 independent / 8 total | |
| DeepSeek V4 Pro (0813) DeepSeek | 2026-08-13 | $1.32 / $3.96off-peak $0.66 / $1.98 | 1M | 5 independent / 9 total | |
| Gemini 3.7 Flash Google DeepMind | 2026-08-13 | $0.75 / $3.75 | 1M | 6 independent / 6 total | |
| Grok 4.6 xAI (SpaceXAI) | 2026-08-12 | $2.00 / $6.00 | 500K | 6 independent / 6 total | |
| Qwen3.8-Max Alibaba | 2026-08-03 | $2.00 / $6.00 | 1M | 6 independent / 9 total | |
| DeepSeek V4 Flash (0731) DeepSeek | 2026-07-31 | $0.44 / $1.32off-peak $0.22 / $0.66 | 1M | 6 independent / 6 total | |
| Claude Opus 5 Anthropic | 2026-07-24 | $5.00 / $25.00 | 1M | 7 independent / 7 total | |
| Kimi K3 Moonshot AI | 2026-07-16 | $3.00 / $15.00 | 1M | 7 independent / 9 total | |
| GPT-5.6 Sol OpenAI | 2026-07-09 | $5.00 / $30.00 | 1M | 7 independent / 7 total | |
| Claude Sonnet 5 Anthropic | 2026-06-30 | $2.00 / $10.00 | 1M | 7 independent / 7 total | |
| GLM-5.2 Zhipu AI (Z.ai) | 2026-06-16 | $1.40 / $4.40 | 1M | Superseded | 7 independent / 9 total |
| Claude Fable 5 Anthropic | 2026-06-09 | $10.00 / $50.00 | 1M | 7 independent / 7 total | |
| Claude Opus 4.8 Anthropic | 2026-05-28 | $5.00 / $25.00 | 1M | Superseded | 8 independent / 9 total |
| Gemini 3.1 Pro Preview Google DeepMind | 2026-02-19 | $2.00 / $12.00 | 1M | 7 independent / 8 total |
Verified = how many of this model’s sourced benchmark scores were run by an independent third party rather than self-reported by the vendor; “—” means no scores are sourced yet.