Anthropic released Mythos 5.2 and Fable 5.2, claiming gains in coding, reasoning and vision, while cutting Fable input pricing from $10 to $8 per million tokens and speeding responses by 30%. It is the lab's first model launch since a METR report suggested it may be approaching recursive self-improvement, and Fable 5.2 now supports zero data retention for enterprise clients.
Zero Data Retention (ZDR)
零数据保留
Enterprise compliance policy under which providers retain no user request or response data.
Curated Intelligence
OpenCode has made its 'stealth model' Cat Jam available for free on the platform. Designed specifically for agentic coding, the model emphasizes high inference speed and a zero data retention policy for privacy. However, no technical specifications, benchmark scores, or architectural details have been disclosed.
Inception AI has added a paid Zero Data Retention (ZDR) endpoint for its decision model, Mercury Decide, on OpenRouter. The new tier offers input pricing at $0.02 per million tokens, free output and cached inputs, and a 66K context window.
OpenCode has announced a one-week free access period for the Step 5 Preview model on its platform. The offering highlights a 1 million token context window, multimodal capabilities, and zero data retention privacy features to lower adoption barriers for developers.
OpenAI has launched the public beta of its new Decisions API, designed for fast text and image evaluations like classification and rating, running approximately ten times faster than the Responses API. Currently supporting only gpt-6-luna with free output tokens, this move addresses the emerging trend of 'decision models' while simultaneously simplifying OpenAI's paid API tier structure.
Anthropic announced Enterprise Frontier Safeguards (EFS), a solution combining Zero Data Retention (ZDR) privacy with advanced misuse detection. By storing data in customer-controlled cloud infrastructure, EFS addresses enterprise security dilemmas and will roll out to platforms like Claude Code and Bedrock this fall.
OpenAI reaffirmed Zero Data Retention for eligible API customers and previewed 'Private Safety Processing' to enforce advanced AI safety without compromising data privacy. The feature remains in a preview stage with no specific technical parameters or launch timeline disclosed.