Anthropic Launches Claude Fable and Mythos 5.2

Anthropic upgraded Claude Fable 5.2 and Claude Mythos 5.2, the same underlying model with different safeguards: Fable is available to all users, while Mythos is reserved for vetted cybersecurity and biomedical organizations. In prerelease testing the models solved open problems in black-hole physics and formalized Fermat's Last Theorem with minimal human guidance. — anthropic.com

Muse Ember Targets the Busywork of AI Agents

Meta released Muse Ember, a small model built for the repetitive work agents do all day, including tool calls and coding subtasks. It delivers more intelligence per dollar than costlier rivals at $0.10 per million input tokens. — research.meta.ai

Google Adds TPU 8tx and 8i Max to Portfolio

Google added TPU 8tx for confidential inference and TPU 8i Max for massive context windows to its TPU line. The 8tx keeps data and model weights protected in hardware during active use; the 8i Max doubles compute and memory per package for agent workloads. — blog.google

Claude Fable Tops the Artificial Analysis Index

Claude Fable scores 65 on the Artificial Analysis Intelligence Index, the highest figure the benchmarking service has recorded. — artificialanalysis.ai

DeepX Tapes Out Chip for 100-Billion-Parameter Edge AI

DeepX taped out the DX-M2, an edge accelerator designed to run generative-AI models with up to 100 billion parameters at low power, letting robots and drones handle such workloads on-device. — eetimes.com

More from this topic