TL;DR
A rumor says Anthropic is close to releasing Claude Opus 4.8 and Sonnet 4.8, but the main market signal has been misread. The cited 70% probability appears tied to a later Polymarket date, while May 31 is priced near 26%, and Anthropic has not announced Claude 4.8.
A rumor that Anthropic is about to release Claude Opus 4.8 and Sonnet 4.8 is spreading among developers, but the market signal behind it appears weaker than the headline claim: the cited 70% probability applies to a later Polymarket date, while a May 31 release is priced at about 26%, and Anthropic has made no official announcement.
The confirmed facts are limited. Claude Opus 4.7 is Anthropic’s current flagship release, launched April 16, 2026, with a listed price of $5 per million input tokens and $25 per million output tokens, according to the source material. Claude Sonnet 4.6, launched February 17, 2026, remains the current Sonnet model at $3 and $15 per million tokens. Both are described as supporting a 1 million token context window.
The main evidence cited by people expecting Claude 4.8 is a March 31 Claude Code source-map string reading sonnet-4-8, found inside what the source material describes as a forbidden model names filter list. That is a real signal, but it is not the same as a release notice, API identifier, model card or benchmark result. There is still no confirmed Anthropic release date for Claude 4.8.
The Polymarket event cited in the rumor asks whether Claude 4.8 will be released by several dates. The source material says the market should be read as a date ladder rather than a single headline number. On that reading, the market was pricing a release as possible soon, more likely by mid-June, and much more likely by late July. The market had about $900 in total volume as of May 27, which makes it a thin sentiment signal rather than strong evidence of Anthropic’s launch plan.
Why It Matters
The distinction matters because developers and companies may delay model choices, product tests or agent rollouts while waiting for a model that has not been confirmed. If the May 31 probability is closer to 26% than 70%, the practical reading is different: a near-term Claude 4.8 release remains possible, but it is not the base case implied by the viral claim.
The rumor also matters because a Sonnet 4.8 release, if it arrives at the current Sonnet price band, could affect coding agents, document-heavy workflows, screenshot analysis and enterprise AI tooling. But those possible gains are still projections. Current production decisions have to rest on Opus 4.7 and Sonnet 4.6, which are available now.
AI/ML Definitive Guide: Architecture, Models, Big Data, Deployment, Open-Source Tools, Cloud Services, MLOps, LLMs, Gen AI
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
The source material frames the current speculation as a combination of three signals: the March 31 sonnet-4-8 string, Anthropic’s recent model release cadence and the Polymarket date ladder. None of those alone confirms a pending release.
One detail complicates the rumor: Anthropic has not released Sonnet 4.7. A direct move from Sonnet 4.6 to Sonnet 4.8 would break the recent pairing pattern between Opus and Sonnet minor versions. That could mean Anthropic plans to resync model names with a dual 4.8 release, or it could mean the leaked string never becomes a public model name.
The source material’s practical takeaway is that teams should keep evaluating the models they can use today. It points to Sonnet 4.6 as the default for many coding-agent tasks and Opus 4.7 for harder work such as large refactors and longer reasoning chains.
“The 70% number is real. But the two have been welded together wrong.”
— Thorsten Meyer AI source material
“May 31 is actually ~26%.”
— Thorsten Meyer AI source material
“No Anthropic announcement, no API id, no model card, no benchmarks for 4.8.”
— Thorsten Meyer AI source material
AWS CERTIFIED GENERATIVE AI DEVELOPER PROFESSIONAL Exam AIP-C01 Complete Study Guide
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
It is not yet clear whether Anthropic will release Claude 4.8, whether both Opus and Sonnet versions would ship together, whether the public model name would use the 4.8 label, or what performance changes the models would bring. Claims about vision gains, SWE-Bench results or enterprise improvements are community extrapolations unless Anthropic publishes model details.
The prediction-market signal is also limited. With about $900 in cited volume, it reflects a small pool of traders and should be treated as a sentiment gauge, not as evidence from Anthropic.
Design Multi-Agent AI Systems Using MCP and A2A: Engineer your own Python-based agentic AI framework with tool use, memory, and multi-agent workflows
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
The next concrete marker is any official Anthropic release note, API model identifier, model card, pricing page update or benchmark publication. Until then, the market’s later date bands and the March source-map string remain signals to watch rather than confirmation of a release.
Choosing the Right AI Engine for Your Organization
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Has Anthropic announced Claude 4.8?
No. Based on the supplied source material, there is no Anthropic announcement, API ID, model card or benchmark release for Claude 4.8.
Does Polymarket show 70% odds for Claude 4.8 by May 31?
No, according to the source material. The 70% figure is tied to a later date on the market’s ladder, while May 31 is cited at about 26%.
What evidence supports the Claude 4.8 rumor?
The main evidence is a March 31 Claude Code source-map string containing sonnet-4-8, plus market pricing that rises across later release dates. Neither confirms a release.
Why is the missing Sonnet 4.7 relevant?
Anthropic has recently paired Opus and Sonnet minor versions. A move from Sonnet 4.6 straight to Sonnet 4.8 would break that pattern, which makes the naming rumor less straightforward.
What should developers use while waiting?
The source material points to Sonnet 4.6 for regular coding-agent work and Opus 4.7 for harder tasks. Claude 4.8 remains unconfirmed.
Source: Thorsten Meyer AI
