Anthropic is preparing a surprise release of its next flagship large language model, Opus 5.5, that could land as early as Tuesday, according to internal artifacts reviewed by 36Kr Europe. The publication broke the story on September 21, 2026 at 08:47 SGT, sending shockwaves through the AI sector less than three weeks after OpenAI debuted GPT-6 and just days after rival labs signaled imminent upgrades. If the rumored launch holds, it would mark one of the most aggressive counterpunches in the recent foundation-model wars.
Opus 5.5: A skipped version line and a leaked codename
According to files shared with 36Kr Europe, Anthropic skipped what would have been an Opus 5.2 checkpoint and jumped straight from Opus 5, released July 24, 2026, to Opus 5.5. The codename “claude-wafer-eap” appeared in closed-beta artifacts, confirming internal maturity beyond the version numbers publicly advertised. The discovery was first made by a developer going by the handle “Lyra,” who traced the unusual build string back to Anthropic’s staging environments. Anthropic has historically followed a steady .0, .1, .2 cadence, and the company appears to have “never planned to follow the regular iteration path” with this release.
Opus 5.5 aimed squarely at GPT-6 Astra
The new model is widely understood inside partner channels as Anthropic’s direct response to OpenAI’s GPT-6, internally branded GPT-6 Astra, which shipped on September 3, 2026. Lyra’s benchmarks, replicated by several independent developers, suggest that Opus 5.5 can outperform gpt-6-sol on a handful of agentic and long-context tasks, a claim that would have seemed improbable only weeks ago. The timing is no coincidence: Anthropic is widely believed to need a blockbuster model before its anticipated IPO, and the company appears to be using Opus 5.5 to restore market confidence after months of investor jitters.
Closed-beta artifacts hint at an out-of-blue launch
Multiple developer accounts describe quiet invitations to a closed-beta program that was not publicly advertised. Rumors now point to a surprise release as soon as Tuesday, September 22, with no formal announcement planned. Anthropic appears comfortable letting the leak do the marketing, a tactic that worked for several stealth launches earlier in the year. The closed-beta artifacts also reference the same “claude-wafer-eap” identifier, suggesting that what is being tested externally is the same build intended for general availability.
Pricing that turns frontier models into commodities
Perhaps the most disruptive element of the leak is the rumored pricing schedule for Opus 5.5. Input tokens are said to be priced at $4 per million, a 20% cut from Opus 5’s $5, while output tokens are reportedly set at $20 per million, down from $25, another 20% reduction. The cache read price, at roughly $0.20 per million tokens, implies an effective discount approaching 90% for workloads that hit Anthropic’s prompt cache, while cache writes are listed at $5 per million. Combined, the numbers point to a strategic bet on volume rather than margin, reframing top-tier large models from luxury goods into daily necessities.
Same-week collisions with OpenAI and Google DeepMind — Why the IPO clock matters
Opus 5.5 continues to define the competitive landscape, and the developments this week underscore how quickly the underlying economics and market structure are shifting. Analysts expect the next month to bring additional disclosure around partnerships, customer commitments, and benchmark performance, all of which will shape how enterprises and consumers evaluate the trade-off between cost, capability, and reliability. The early signal points to a market in which opus 5.5 sets the new baseline rather than the ceiling, with rivals forced to match on performance or price to remain relevant.
Opus 5.5 has become the defining story of the ai beat this week, and the broader industry is responding. Analysts at major research desks have begun updating their forecasts, and customer commitments are likely to follow within days. The pace of opus 5.5’s rollout sets a benchmark that competitors will struggle to match on cost without sacrificing capability, and capability without sacrificing cost. Expect additional disclosures over the coming weeks as partners and customers publish their own evaluations and reference deployments.
Looking ahead, the question for opus 5.5 is whether the early advantage holds as rivals regroup. The combination of timing, scale, and a well-articulated roadmap makes this a hard story to dismiss, but history shows that first-mover status in this market rarely lasts without sustained execution. Watch for follow-on customer announcements, third-party benchmarks, and pricing moves from incumbents over the next 30 to 60 days as the real signal of whether opus 5.5 reshapes the competitive landscape or merely resets the bar.
Looking ahead, the question for Opus 5.5 is whether the early advantage holds as rivals regroup. The combination of timing, scale, and a well-articulated roadmap makes this a hard story to dismiss, but history shows that first-mover status in this market rarely lasts without sustained execution. Watch for follow-on customer announcements, third-party benchmarks, and pricing moves from incumbents over the next 30 to 60 days as the real signal of whether Opus 5.5 reshapes the competitive landscape or merely resets the bar.