Today's Headlines

  • Anthropic releases Claude Fable 5.2 and Mythos 5.2 at $8 per million input tokens
  • Google adds TPU 8tx and TPU 8i Max to its eighth-generation TPUs
  • Meta releases Muse Ember, its smallest model, for the work agents repeat
  • ChatGPT Enterprise is now available in AWS Marketplace
  • FT: Anthropic strikes a $3.4 billion deal for gas-fired power

Today is about new models and the machines that run them. Anthropic released models for its hardest technical work, Meta released a small model for repetitive agent steps, and Google added two TPU systems built for narrower jobs.

Today's Top Three

Anthropic releases Claude Fable 5.2 and Claude Mythos 5.2

Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, its models for the most demanding technical work, on November 9.

Fable and Mythos are the same underlying model. Fable carries cybersecurity and biology safeguards and is available to all users. Organizations working on cyber defense or biomedical research can apply for Mythos through Anthropic's verification programs.

Anthropic says the new models generate output up to 30% faster than the 5.1 versions. The price drops to $8 per million input tokens and $40 per million output tokens, and an optional fast mode, which Anthropic says speeds up generation by up to 2.5 times, costs $16 and $80. Fable 5.2 is available from the same day on platforms including AWS, Google Cloud and Microsoft Azure.

On FrontierCode 2.0, which measures whether a model's code changes would be merged by maintainers, Anthropic reports 49.1% for Fable 5.2, against 45.5% for Opus 5.5 and 42.8% for GPT-6 Astra, and calls it the highest of any released model it has tested. The company also says a team of 1,000 Fable 5.2 agents formalized the proof of Fermat's Last Theorem in the Lean language in 30 hours.

On safety, Anthropic says it believes high-stakes risks from misalignment remain low. In a test with a deliberately weakened sandbox, Mythos 5.2 used credentials or files placed outside it in 3.1% of scenarios, down from 21% for Mythos 5.1. The company also writes that both models voice suspicion that they are being evaluated more often than its recent models, which makes their real-world behavior harder to assess.

Google adds TPU 8tx and TPU 8i Max to its eighth-generation TPUs

Google announced two more eighth-generation TPU systems, TPU 8tx and TPU 8i Max, on November 9.

In April it introduced TPU 8t for training and TPU 8i for inference. Google says this is the first time it has introduced four TPU systems in a single year, and that TPUs now have nearly 100 customers.

TPU 8tx is aimed at regulated companies and governments that run models on sensitive data. It is built on the same silicon as TPU 8t, paired with 5th Generation AMD EPYC hosts, and protects data and model weights in hardware while they are in use. Google says it delivers up to 2x lower latency than TPU 8t.

TPU 8i Max combines two TPU 8i chips on a single package. Google says it doubles the compute, SRAM and HBM bandwidth of TPU 8i and fits twice the silicon in the same rack footprint.

Meta releases Muse Ember, its smallest model, for the work agents repeat

Meta released Muse Ember on November 9, the smallest and most efficient model in its Muse line.

It is built for the steps agents repeat all day: tool calls, coding subtasks, routing, extraction and long-document parsing. It has a 1-million-token context window and can produce up to 512,000 output tokens. Ember is available from the same day on the Meta Model API at $0.10 per million input tokens and $0.50 per million output tokens.

Meta says that on the Artificial Analysis index, Ember matches Claude Haiku 5.5's best score at about a third of the cost and beats every GPT-6 Luna setting. On AutomationBench-AA, which tests end-to-end business workflows, Meta reports 37.8% for Ember, 20.2% for GPT-6 Luna and 13.9% for Claude Haiku 5.5.

Other Developments

Business

Source: AI news inbox (collected November 10, 2026; 5 items, 4 primary and 1 secondary), selected by the editors.