AI Knowledge Cutoff & Release Date Lookup
Pick any major LLM and get two dates straight from the provider's model card — the training-data knowledge cutoff and the public release date — plus how stale its knowledge is today. No signup, no API, sources cited below.
How it works
This is a lookup, not a calculator: the dates themselves are editorial. Each model's knowledge cutoff and release month are transcribed verbatim from the provider's own model card or model documentation, and every row links back to the exact source it was read from. Nothing is inferred — where a vendor does not publish a cutoff at all (Mistral AI, xAI, DeepSeek), the tool shows “—” instead of a guess.
The only thing computed here is knowledge age, and it is fully deterministic calendar arithmetic:
- Each cutoff is stored as a normalized year-month
YYYY-MM. Where a provider states only a season or quarter, the conservative latest month of that period is used and footnoted on the model. - On load the page reads today's year-month as the reference point
now(this page is rendered per request, and the displayed verified date makes the snapshot explicit). - Knowledge age in months is
(now.year − cutoff.year) × 12 + (now.month − cutoff.month), clamped at zero so a future-dated cutoff can never show a negative age. - The freshness chip is presentational only: ≤6 months is fresh, 7–18 is aging, and over 18 is stale. These thresholds describe recency of training data, not model quality.
The same subtraction is implemented twice — once directly and once via an absolute month index (year × 12 + month) — and the two are required to agree, so an arithmetic regression would be caught rather than shipped. Because release dates lag cutoffs by months, the tool also shows how long ago each model was released, to make the gap between “trained” and “available” visible at a glance.
Worked examples
Reference point for all three: June 2026 (the 2026-06-13 snapshot month). The live tool recomputes against the actual current month.
Frequently asked questions
Sources & references
Each model row links to the specific provider page it was transcribed from. The provider model-card hubs:
- OpenAI — model docs (GPT-5, GPT-4o, GPT-4.1, GPT-4 Turbo)
- Anthropic — Claude model overview (Opus & Sonnet 4.x, 3.7, 3.5)
- Google — Gemini model docs (2.5, 2.0)
- Meta — Llama model cards (Llama 4, 3.1)
- Mistral AI — model docs
- xAI — Grok model docs
- DeepSeek — model cards on Hugging Face
Dates were last cross-checked against these sources on 2026-06-13. This is a cited snapshot, not a live feed — model cards change, so the verified date above always tells you how current the data is. Spotted a date that has moved? Email me and I'll update the row.
Related tools
Comments & feedback
Spotted a bug or want an improvement? Tell us — our team reviews every comment, and good ideas get built. Comments are public and anonymous.
Found an out-of-date row, a model we're missing, or an edge case?
Email me at [email protected] — most fixes ship within 24 hours.