Live forecast · 17 models tracked
When is the next major AI model expected to be released?
Meta's "Watermelon" model · market-implied median
Prediction markets currently put the Meta's "Watermelon" model release around Oct 13, 2026, with a middle-50% window from Oct 12, 2026 to Oct 14, 2026. Markets imply a 90% chance of landing within a week of that date, and 95% within a month. The estimate is based on 3 active deadline contracts with $96K of traded volume.
Every tracked model
Estimated AI model release dates
| Model | Lab | Expected release | Within a week |
|---|---|---|---|
| Next Google Gemini Pro Model | Google DeepMind | About Feb 22, 2027 | 3.1% |
| Gemini 4.0 | Google DeepMind | Oct 15, 2026 | 80% |
| Next Fable Model (5.2+) | Anthropic | Oct 20, 2026 | 70% |
| GPT Astra 6.1+ | OpenAI | Nov 30, 2026 | 17% |
| SSI public model | Frontier lab | Dec 15, 2026 | 5.1% |
| Meta's "Watermelon" model | Meta | Oct 14, 2026 | 90% |
| Next OpenAI GPT Terra (5.7+) | OpenAI | About Dec 26, 2026 | 4.0% |
| Gemini Argon | Google DeepMind | Oct 14, 2026 | 72% |
| Claude 6 | Anthropic | May 5, 2027 | 5.9% |
| Next Muse Spark (1.4+) | Meta | Oct 25, 2026 | 42% |
| Next Z.ai GLM (5.4+) | Frontier lab | Oct 29, 2026 | 32% |
| Next DeepSeek Pro | DeepSeek | About Jan 17, 2027 | 4.3% |
| Kimi K4 | Moonshot AI | Not yet priced | — |
| Grok 5 | xAI | About Jan 7, 2027 | 16% |
| GPT-7 | OpenAI | Jul 29, 2027 | 4.0% |
| Next Google Gemini Flash-Lite Model (3.6+) | Google DeepMind | Nov 24, 2026 | 26% |
| Next Alibaba Qwen Max (3.9+) | Alibaba | Not yet priced | — |
Open any model for its probability curve, deadline evidence, and recorded forecast history. Machine-readable copies live at /llms.txt, /api/forecasts, and /feed.xml.
Common questions
About the Meta's "Watermelon" model forecast
When is Meta's "Watermelon" model expected to be released?
Prediction markets currently put the Meta's "Watermelon" model release around Oct 13, 2026, with a middle-50% window from Oct 12, 2026 to Oct 14, 2026. Markets imply a 90% chance of landing within a week of that date, and 95% within a month. The estimate is based on 3 active deadline contracts with $96K of traded volume.
How precise is the Meta's "Watermelon" model release date forecast?
Markets place 90% of the probability within a week of the median date and 95% within a month. A date with only a few percent inside a week should be read as a month, not a day.
Where does the Meta's "Watermelon" model release forecast come from?
Each public "released by <date>?" contract on Polymarket gives one point on a cumulative probability curve. Release Oracle sorts those points, corrects small inconsistencies, then interpolates the 50% crossing for the headline date and the 25% and 75% crossings for its uncertainty window. Forecasts are market signals, not lab announcements, and can be wrong.