On Thursday, Anthropic released Opus 4.8, the newest version of its most advanced publicly available model. The model is available everywhere, with standard pricing at the same level as the previous Opus release - because nothing says "innovation" like charging the same amount for a slightly less disappointing product.

The new model comes just 41 days after Opus 4.7 was released, a much faster upgrade cycle than normal for Anthropic. (The most recent Sonnet and Haiku models are three and seven months old, respectively.) The fast turnaround may have something to do with the chilly reception to Opus 4.7, which some users found disappointing - which is tech-speak for "everyone rolled their eyes."

That interval has also seen significant new releases for OpenAI’s Codex and Google’s Gemini Flash model, increasing the pressure on Anthropic to keep pace. Nothing like a little sibling rivalry to get the code flowing.

Opus 4.8 comes with the expected best-in-class benchmark results, but there’s also particular attention to how the model manages bad or uncertain data. In the launch post, Anthropic’s early testers found that the new model is “more likely to flag uncertainties about its work and less likely to make unsupported claims.” In other words, it's finally learning to say "I don't know" instead of confidently hallucinating.

Echoing this point, a testimonial from Bridgewater Associates said the biggest difference in the upgrade was “Opus 4.8’s tendency to proactively flag issues with the inputs and outputs of an analysis, something other models routinely missed and left to the users to catch.” So it's basically the office colleague who points out the spreadsheet errors before the boss sees them.

Together with the new model, Anthropic launched a feature called Dynamic Workflows, which will be available in research preview. The system is designed to help larger models like Opus manage complex tasks across hundreds of parallel subagents. Because one AI managing a thousand tasks wasn't ambitious enough.

“Claude Code alongside Opus 4.8 can now carry out codebase-scale migrations across hundreds of thousands of lines of code from kickoff to merge, with the existing test suite as its bar,” the post explains. That's a lot of code for a model that just learned to admit it's confused.

Anthropic is still holding back its most advanced Mythos model after a tentative preview last month raised cybersecurity concerns. However, the company hinted in today’s Opus release that the Mythos preview period might soon end, once necessary safeguards are complete.

“We’re making swift progress on developing these safeguards and expect to be able to bring Mythos-class models to all our customers in the coming weeks,” the company wrote. Translation: "We've almost figured out how to stop it from accidentally launching nukes."