What is Claude Opus?

Nathan Labenz on why thin-margin AI wrapper companies struggle to compete on security-critical deployments.

“When it's security critical, I think people will still pay up for the very best model. A company running thin margins on top of Opus is going to struggle to say, 'No, don't use that. Use us.'”

Closed-weight labs like Anthropic also use distillation internally. Claude Sonnet is described as a partially distilled version of Claude Opus.

“The closed weight model labs distill models too like you know sonnet is a partially distilled version of opus.”

Cursor is teasing a new frontier-scale model trained from scratch, the same size as Claude Opus and GPT-5, using 10 to 20x more compute than its previous Composer model, designed to be generally intelligent rather than coding-only.

“New cursor model being teased at compile. Same size as Claude Opus and GPT55. Trained from scratch. No more Kimmy Base. 10 to 20x more compute versus composer generally intelligent not just coding releases in the next couple of weeks.”

Open-source model GLM2 is approximately 90% cheaper to use than Claude Opus

“It is 90% cheaper to use these products.”

Cursor's forthcoming model uses 10 to 20x more compute than its prior Composer model, matching the scale of Claude Opus and GPT-5.

“New cursor model being teased at compile. Same size as Claude Opus and GPT55. Trained from scratch. No more Kimmy Base. 10 to 20x more compute versus composer generally intelligent not just coding releases in the next couple of weeks.”

A cost-efficient emerging workflow: use an expensive frontier model (Claude Opus or GPT) once to generate a project taste file, then use cheap models for all subsequent development guided by that file.

“A lot of people what they're doing is they're building one project with a really high quality LLM like Opus or GPT They're building a taste file and then you know, super cheap models to continuously build on that more, you know, project with that taste file.”

Claude Fable low-effort tier reportedly produces better output than Claude Opus high-effort tier, inverting the expected quality hierarchy.

“In fact, Fable low is better than Opus high.”

Greg Isenberg on Claude Fable tier performance relative to Claude Opus.

“In fact, Fable low is better than Opus high.”