Meta's Watermelon frontier large language model remains in training as of mid-September 2026, with trader sentiment shaped by an August internal-document report targeting an October release alongside the Hatch AI agent platform. Alexandr Wang stated in a July town hall that the model matches OpenAI's GPT-5.5 on unnamed benchmarks while using roughly 10x the compute of Muse Spark (codenamed Avocado), though no external evaluations or model cards exist. Mark Zuckerberg publicly teased "Next up 🍉" on September 2, and Wang followed with confirmation that development stays on track for strong competitiveness. Key swing factors include whether Meta meets the October window amid heavy infrastructure spending, integrates the model into consumer products, or encounters further delays typical of frontier training runs.
基於Polymarket數據的AI實驗性摘要。這不是交易建議，也不影響該市場的結算方式。 · 更新於$45,309 交易量
2026-10-31
9 月 30 日
11%
10 月 31 日
54%
11 月 30 日
89%
$45,309 交易量
9 月 30 日
$16,308 交易量
11%
10 月 31 日
$18,912 交易量
54%
11 月 30 日
$10,089 交易量
89%
On September 2, 2026, alongside the launch of Muse Spark 1.3, Meta announced a roadmap "including bigger models" (https://research.meta.ai/blog/introducing-muse-spark-1-3). Credible reporting has identified Meta's next-generation frontier model by the internal codename "Watermelon."
This market will resolve to "Yes" if Meta releases "Watermelon" or a model confirmed to be the model referenced above, and that model is made available to the general public by the listed date (ET). Otherwise, this market will resolve to "No".
A qualifying model must be named "Watermelon" or be identified, by Meta or by a consensus of credible reporting, as the model internally codenamed "Watermelon," regardless of the name under which it is ultimately released.
A qualifying model must be launched and publicly accessible, including via open beta or open rolling waitlist signups. A closed beta or any form of private access will not suffice. The release must either be clearly defined and publicly announced by Meta as accessible to the general public, or otherwise be made publicly accessible and explicitly labeled on the company's official website. Labeling errors, placeholder text, or version names displayed on the website that do not correspond to a model that is actually accessible to the general public will not qualify.
The primary resolution source for this market will be official information from Meta, with additional verification from a consensus of credible reporting.Meta's Watermelon frontier large language model remains in training as of mid-September 2026, with trader sentiment shaped by an August internal-document report targeting an October release alongside the Hatch AI agent platform. Alexandr Wang stated in a July town hall that the model matches OpenAI's GPT-5.5 on unnamed benchmarks while using roughly 10x the compute of Muse Spark (codenamed Avocado), though no external evaluations or model cards exist. Mark Zuckerberg publicly teased "Next up 🍉" on September 2, and Wang followed with confirmation that development stays on track for strong competitiveness. Key swing factors include whether Meta meets the October window amid heavy infrastructure spending, integrates the model into consumer products, or encounters further delays typical of frontier training runs.
On September 2, 2026, alongside the launch of Muse Spark 1.3, Meta announced a roadmap "including bigger models" (https://research.meta.ai/blog/introducing-muse-spark-1-3). Credible reporting has identified Meta's next-generation frontier model by the internal codename "Watermelon."
This market will resolve to "Yes" if Meta releases "Watermelon" or a model confirmed to be the model referenced above, and that model is made available to the general public by the listed date (ET). Otherwise, this market will resolve to "No".
A qualifying model must be named "Watermelon" or be identified, by Meta or by a consensus of credible reporting, as the model internally codenamed "Watermelon," regardless of the name under which it is ultimately released.
A qualifying model must be launched and publicly accessible, including via open beta or open rolling waitlist signups. A closed beta or any form of private access will not suffice. The release must either be clearly defined and publicly announced by Meta as accessible to the general public, or otherwise be made publicly accessible and explicitly labeled on the company's official website. Labeling errors, placeholder text, or version names displayed on the website that do not correspond to a model that is actually accessible to the general public will not qualify.
The primary resolution source for this market will be official information from Meta, with additional verification from a consensus of credible reporting.
This market will resolve to "Yes" if Meta releases "Watermelon" or a model confirmed to be the model referenced above, and that model is made available to the general public by the listed date (ET). Otherwise, this market will resolve to "No".
A qualifying model must be named "Watermelon" or be identified, by Meta or by a consensus of credible reporting, as the model internally codenamed "Watermelon," regardless of the name under which it is ultimately released.
A qualifying model must be launched and publicly accessible, including via open beta or open rolling waitlist signups. A closed beta or any form of private access will not suffice. The release must either be clearly defined and publicly announced by Meta as accessible to the general public, or otherwise be made publicly accessible and explicitly labeled on the company's official website. Labeling errors, placeholder text, or version names displayed on the website that do not correspond to a model that is actually accessible to the general public will not qualify.
The primary resolution source for this market will be official information from Meta, with additional verification from a consensus of credible reporting.
市場開放時間： Sep 4, 2026, 10:35 AM ET
交易量
$45,309結束日期
2026-11-30市場開放時間
Sep 4, 2026, 10:35 AM ETOn September 2, 2026, alongside the launch of Muse Spark 1.3, Meta announced a roadmap "including bigger models" (https://research.meta.ai/blog/introducing-muse-spark-1-3). Credible reporting has identified Meta's next-generation frontier model by the internal codename "Watermelon."
This market will resolve to "Yes" if Meta releases "Watermelon" or a model confirmed to be the model referenced above, and that model is made available to the general public by the listed date (ET). Otherwise, this market will resolve to "No".
A qualifying model must be named "Watermelon" or be identified, by Meta or by a consensus of credible reporting, as the model internally codenamed "Watermelon," regardless of the name under which it is ultimately released.
A qualifying model must be launched and publicly accessible, including via open beta or open rolling waitlist signups. A closed beta or any form of private access will not suffice. The release must either be clearly defined and publicly announced by Meta as accessible to the general public, or otherwise be made publicly accessible and explicitly labeled on the company's official website. Labeling errors, placeholder text, or version names displayed on the website that do not correspond to a model that is actually accessible to the general public will not qualify.
The primary resolution source for this market will be official information from Meta, with additional verification from a consensus of credible reporting.Meta's Watermelon frontier large language model remains in training as of mid-September 2026, with trader sentiment shaped by an August internal-document report targeting an October release alongside the Hatch AI agent platform. Alexandr Wang stated in a July town hall that the model matches OpenAI's GPT-5.5 on unnamed benchmarks while using roughly 10x the compute of Muse Spark (codenamed Avocado), though no external evaluations or model cards exist. Mark Zuckerberg publicly teased "Next up 🍉" on September 2, and Wang followed with confirmation that development stays on track for strong competitiveness. Key swing factors include whether Meta meets the October window amid heavy infrastructure spending, integrates the model into consumer products, or encounters further delays typical of frontier training runs.
On September 2, 2026, alongside the launch of Muse Spark 1.3, Meta announced a roadmap "including bigger models" (https://research.meta.ai/blog/introducing-muse-spark-1-3). Credible reporting has identified Meta's next-generation frontier model by the internal codename "Watermelon."
This market will resolve to "Yes" if Meta releases "Watermelon" or a model confirmed to be the model referenced above, and that model is made available to the general public by the listed date (ET). Otherwise, this market will resolve to "No".
A qualifying model must be named "Watermelon" or be identified, by Meta or by a consensus of credible reporting, as the model internally codenamed "Watermelon," regardless of the name under which it is ultimately released.
A qualifying model must be launched and publicly accessible, including via open beta or open rolling waitlist signups. A closed beta or any form of private access will not suffice. The release must either be clearly defined and publicly announced by Meta as accessible to the general public, or otherwise be made publicly accessible and explicitly labeled on the company's official website. Labeling errors, placeholder text, or version names displayed on the website that do not correspond to a model that is actually accessible to the general public will not qualify.
The primary resolution source for this market will be official information from Meta, with additional verification from a consensus of credible reporting.
This market will resolve to "Yes" if Meta releases "Watermelon" or a model confirmed to be the model referenced above, and that model is made available to the general public by the listed date (ET). Otherwise, this market will resolve to "No".
A qualifying model must be named "Watermelon" or be identified, by Meta or by a consensus of credible reporting, as the model internally codenamed "Watermelon," regardless of the name under which it is ultimately released.
A qualifying model must be launched and publicly accessible, including via open beta or open rolling waitlist signups. A closed beta or any form of private access will not suffice. The release must either be clearly defined and publicly announced by Meta as accessible to the general public, or otherwise be made publicly accessible and explicitly labeled on the company's official website. Labeling errors, placeholder text, or version names displayed on the website that do not correspond to a model that is actually accessible to the general public will not qualify.
The primary resolution source for this market will be official information from Meta, with additional verification from a consensus of credible reporting.
交易量
$45,309結束日期
2026-11-30市場開放時間
Sep 4, 2026, 10:35 AM ETMeta's Watermelon frontier large language model remains in training as of mid-September 2026, with trader sentiment shaped by an August internal-document report targeting an October release alongside the Hatch AI agent platform. Alexandr Wang stated in a July town hall that the model matches OpenAI's GPT-5.5 on unnamed benchmarks while using roughly 10x the compute of Muse Spark (codenamed Avocado), though no external evaluations or model cards exist. Mark Zuckerberg publicly teased "Next up 🍉" on September 2, and Wang followed with confirmation that development stays on track for strong competitiveness. Key swing factors include whether Meta meets the October window amid heavy infrastructure spending, integrates the model into consumer products, or encounters further delays typical of frontier training runs.
基於Polymarket數據的AI實驗性摘要。這不是交易建議，也不影響該市場的結算方式。 · 更新於
警惕外部連結哦。
警惕外部連結哦。
Frequently Asked Questions