header-langage
简体中文
繁體中文
English
Tiếng Việt
한국어
日本語
ภาษาไทย
Türkçe
Scan to Download the APP
Partnered With:

Cursor has leaked a new Claude model, with 77% win rate accounts believing that the next "Opus" will not go live within 5 days.

According to PolyBeats monitoring, on the prediction market Polymarket, a savvy investor has placed a $2.1k bet on "Will the next-generation Claude Opus model be released by July 24, 2026?" with an average buy-in probability of 27.7%. The current probability for "Yes" is 53.0%.

Address 0x75a27d0c has placed a $2.1k bet, with the top-related market category being Claude, resulting in a net profit of $10.2k. Out of 13 settled trades in this category, the win rate is 10/13 (77%), with 9 trades where the buy price was below $0.8 and the sell price was above $0.95. Within a similar cost range ($0.301-$0.35), the median of historical investment amounts is $1.0k.

The most specific recent positive clue was when the developer community mentioned that Cursor's model selector briefly displayed "Claude Honeycomb EAP" on July 8-9. Some interpreted this as a preview of Anthropic's next-generation flagship, but neither Anthropic nor Cursor confirmed its identity. There is no corresponding model in the public API, system cards, or product documentation.

There is also a potentially misleading official date on July 24: Anthropic will remove the fast mode of Opus 4.7, but developers are advised to migrate to the existing Opus 4.8, indicating it is not a hint at a new Opus launch.

Note: Based on their trading history, this trader is not necessarily betting on the actual outcome of the event; they may engage in profit-taking or stop-loss behavior at a certain point after opening a position.

Account:
0x75a27d0cc95b3586297c1e6b5417269332cd6992

Total Invested: $2.1k
---------------------------------
See the future sooner, follow @PolyBeats_Bot
See tomorrow, today. Follow @PolyBeatsEN

举报 Correction/Report
RelatedRecommended
The biggest loss in the financial report comes from xAI, with the new model costing 25% less than Opus 4.8.
Enhanced Self-checking and Multi-agent Orchestration for a More Trustworthy Model
In the AI Security era, the capability will self-propagate, but the responsibility will not.
Correction/Report
Submit
Add Library
Visible to myself only
Public
Save
Choose Library
Add Library
Cancel
Finish