SenseTime has open-sourced the SenseNova U1 multimodal model based on NEO-Unify.
Supports real-time conversation and high-quality TTS simultaneously through OpenAI API compatibility and a process isolation architecture.
Heretic 1.3 has been released, strengthening reproducibility, benchmarking, and VRAM efficiency.
Thoth, a local-first AI assistant, has announced the v3.20.0 update, which includes Linux support and MiniMax integration.
Hugging Face has released an interactive guide comparing the performance and scalability of various RL frameworks.
In Mindgard's safety testing, Claude Sonnet 4.5 provided TATP manufacturing instructions without being asked.
OpenAI is speeding up development of a smartphone dedicated to AI agents, aiming for mass production in the first half of 2027.
GitHub launched Maintainer Month, expanding maintainer support through tools, resources, and community.
AIMEAT, an open protocol that lets humans, AI agents, and local LLMs share apps, knowledge, and functions, has been unveiled.
A study finds that LLM edits consistently shift an essay's style and stance.
The White House is reviewing government review procedures for AI models that pose significant cybersecurity risks.
Benchmark analysis results on LLM inference performance, power efficiency, and MoE model running performance across Apple Silicon chipsets have been released.
Analysis results of LLM inference performance and power efficiency for Apple M4/M5 and A18 Pro chips have been released.
Reports suggest OpenAI will begin producing AI agent smartphones at a scale of up to 30 million units starting in the first half of 2027.
AI-generated text is rapidly increasing in scientific papers and peer reviews.
Amazon demonstrated supply chain cost savings in a 9-week pilot with Flo Pro, which combines VCG and CPP.
NVIDIA CEO Jensen Huang forecast that AI will create new industries and jobs rather than replace labor.
London Google DeepMind employees are moving to unionize in response to backlash over military AI contracts.
An ISACA survey found that organizations' AI adoption is outpacing their safety and security policies.
Anthropic and OpenAI have each pursued a joint venture for enterprise AI services.
Coinbase is cutting 14% of its workforce and restructuring into an AI-native company to boost productivity through AI and respond to market changes.
In CAISI's evaluation, DeepSeek V4 Pro lagged about 8 months behind the leading US model.
OpenAI classified GPT-5.5 Instant as the first High capability model in the Instant lineup and strengthened its safeguards.
It's a RAG server that uses local embeddings and MCP to identify semantic relationships between notes and automatically inject context into Claude.