French-language name; a family of AI language models by Anthropic
Beijing-based Moonshot AI has released Kimi K3, a 2.8‑trillion‑parameter open‑weight model that has outperformed leading U.S. systems on blind coding benchmarks, is priced substantially below comparable U.S. offerings, and is scheduled to publish its model weights on July 27 so developers can download, modify and self‑host the model.
OpenRouter and Concentrate AI are expanding AI-model routing, driving demand for cheaper models and cost controls as Anthropic and OpenAI race to go public. The landscape shifts as buyers seek value and governments debate how to share benefits from AI progress.
A coalition of U.S. state attorneys general has subpoenaed OpenAI for internal documents on advertising, user engagement, handling of health and consumer data, and protections for minors and seniors. OpenAI has said it will "engage constructively," highlighted new safeguards in ChatGPT and is cooperating with investigators while facing related lawsuits and regulatory pressure.
Snap has unveiled Specs, its consumer AR glasses, priced at $2,195 with a $200 refundable deposit. The device aims to merge AI with real-world computing, featuring EyeConnect for shared experiences and on-device AI. Market reaction remains cautious as investors scrutinize affordability and the competitive hardware landscape.
Meta has announced a new line of smart glasses with AI-assisted features, priced at $299 and developed with EssilorLuxottica. The glasses lack Ray-Ban/Oakley branding, and Meta is positioning them as fashion-forward wearables. Competitors like Snap and Google are racing to release their own AI eyewear.
The US Department of Commerce has lifted export controls on Anthropic's Claude Fable 5 and Mythos 5, and Anthropic has begun restoring access. Mythos 5 has been cleared for a vetted group of US organisations; Fable 5 — redesigned with stronger safeguards — is being redeployed more broadly after testing and coordination with government officials.
Anthropic has introduced Claude Science, a workspace that links Claude models to 60+ databases and specialized toolkits to automate life-sciences tasks. The platform includes a project-management workflow, sub-assistants, and a fact-checker to improve reproducibility. Early users report rapid genome browser creation and computational review pipelines.
Anthropic has redeployed Fable 5 globally with stronger safeguards after government export curbs, making the model available again to most customers while routing risky queries to weaker models. The move follows earlier shutdowns and a government order; industry and policymakers are watching how future frontier models will be released.
Anthropic has released Claude Reflect to web and mobile for users with memory enabled, allowing them to review 1, 3, 6 or 12 months of Claude activity, usage patterns, and task types. Prompts about mindful AI use and optional quiet hours accompany the rollout, which is in beta for Free, Pro, and Max tiers.
Enterprises are shifting budgeting to value over token costs as AI costs rise. Leaders argue for metrics based on useful work and task-level economics, with pilots testing open-weight and sovereign models.
Enterprise leaders confront rising token costs as tokenmaxxing shifts from hype to pragmatism. Leaders say tokenomics must guide responsible exploration of agentic AI, with a focus on cost efficiency and value.