Odds on Anthropic’s next Opus model surges despite calls to slow AI development
Photo by Solen Feyissa on Unsplash.
- Polymarket traders price a September 24 launch of Anthropic’s next Claude Opus model at 83%.
- Anthropic says AI now drives 26% of its R&D work and more than 80% of merged code.
- The prediction market’s launch bets come as Anthropic CEO Dario Amodei and former researchers urge the AI industry to slow frontier-model development.
Prediction-market traders are betting that Anthropic will ship its next Claude Opus model within days.
This is coming at a time when Anthropic’s own chief executive and a run of departing researchers press the industry to ease off the accelerator.
Traders price in a late-September Opus
On Polymarket’s “Next Claude Opus released by…?” market, the contract for a public release by September 24 was trading around 83 cents on the dollar as of Monday, September 21, with each cent roughly corresponding to a percentage-point chance.
The trend so far is that the farther the date, the higher the confidence, which is currently 94% for September 30, about 98% by October 31, and above 99% by year-end.
The market, which opened on September 15, has drawn roughly $240,925 in volume.
A qualifying model has to be a publicly accessible release explicitly named Opus, with an open beta or rolling waitlist counting. Other models, such as Sonnet, Haiku, Fable and Mythos, do not qualify.
Anthropic has not released any official announcement to confirm any release date or version number. Polymarket’s own market page ties the late-September consensus to reported stealth testing of a model codenamed claude-wafer-eap, possibly Claude Opus 5.5. However, it is still an unverified speculation.
Four Claude 5 models in under two months
Anthropic has been busy with releases this year, and it may be the reason why traders are opening markets on future releases. The company released Claude Opus 5 on July 24, and it was the fourth Claude 5 model in less than two months.
Anthropic then released Fable 5.1 on September 1.
The pace of releases from Anthropic is not unique to it, as other AI labs have been keeping that pace as well.
A Nikkei survey of five leading US labs and four Chinese firms, including Alibaba and Z.ai, found the average gap between high-performance model updates has collapsed to 44 days for the April-to-September 2026 window. This is down from 125 days over the roughly three years through March 2026.
AI is now writing most of Anthropic’s code
Per Anthropic, the reason for the speed-up is that AI is increasingly building AI. The company published data on September 17 that shows that Claude led about 26% of its research and development work as of August. This is up from close to zero in February, with more than 90% of that work involving AI in some form.
Claude also authored more than 80% of the code merged into its codebase as of May 2026. A typical engineer was merging roughly eight times as much code per day in the second quarter of 2026 as in 2024.
The Anthropic Institute sees this as a step toward recursive self-improvement, while cautioning that the milestone has not been reached and is “not inevitable.”
OpenAI reported a parallel trend, with AI agents logging 3.1 times the working hours of its human researchers in August.
The people asking Anthropic to slow down
The pace of model releases has also ruffled some feathers, especially those who are concerned about the capabilities of these models and how the labs are training the models.
Anthropic has been at the forefront of most of the criticisms. Its CEO Dario Amodei published an essay titled “We Must Pace the Frontier” on September 12, a call that drew public backing from OpenAI’s Sam Altman and Elon Musk.
Amodei’s post came a few days after former Anthropic employee Jacob Coxon quit on September 9 and went on to say that there is “a strong chance that we could all die in the immediate future.”
His warning is similar to Bilal Chughtai’s, who left Google DeepMind’s safety team in July.
Microsoft AI’s CEO Mustafa Suleyman added his own critique on September 16, where he called out how Anthropic is training its Claude models. Suleyman said that it is a dangerous mistake and is pushing for industry norms that will keep humans in control.
President Trump does not seem to share the same position with the industry, as he has dismissed the doom talk, saying that his priority is beating China in the AI race.
The smartest crypto minds already read our newsletter. Want in? Join them.
FAQs
When could Anthropic release its next Claude Opus model?
Polymarket traders were pricing about an 83% chance of a public release by September 24 and roughly 94% by September 30, but Anthropic has not confirmed any date or version, and the reported testing of a "claude-wafer-eap" model is unverified.
What was Anthropic's most recent Opus release?
Anthropic launched Claude Opus 5 on July 24 at $5 per million input tokens and $25 per million output tokens, which Axios described as the company's fourth Claude 5 model in under two months.
Why are people calling on Anthropic to slow AI development?
CEO Dario Amodei published a "We Must Pace the Frontier" essay on September 12, and figures including departed Anthropic researcher Jacob Coxon and Microsoft AI's Mustafa Suleyman have warned that the speed of AI development, now partly automated by AI itself, poses serious control and safety risks.
Hannah Collymore
Hannah is a writer and editor with nearly a decade of blog writing and event reporting experience in the crypto space. At Cryptopolitan, Hannah contributes to the news page, reporting and analyzing the latest developments in DeFi, RWA, crypto regulation, AI and frontier tech industries. She graduated from Arcadia university with a degree in Business Administration.
