Databricks will make GLM 5.2 its default coding model after internal tests found Opus-like performance at lower per-task cost on real codebase work.
OpenAI has launched the GPT-5.6 model family with Sol, Terra, Luna and a new multi-agent ultra mode, aiming to improve AI performance per dollar.
OpenAI says about 30% of SWE-Bench Pro tasks may be broken, raising new doubts about how the AI industry measures coding models.
Nous Research has released NousCoder-14B, an Apache 2.0 coding model with an open training stack, as demand for AI programming tools intensifies.
Nvidia CEO Jensen Huang says more software work is moving from Python coding to AI agents, underscoring a broader shift in how teams build AI products.
Tencent sold about $1.5 billion of its Kuaishou stake, a move investors are reading as part of its faster capital shift toward AI bets.
First Recon AI has introduced an AI Security Runtime focused on audit-ready evidence, reflecting growing enterprise demand for tighter AI governance.
OpenAI has introduced GPT-Live voice models for ChatGPT, betting that full-duplex speech can make AI assistants more useful for live work.
Nvidia and LangChain introduced NemoClaw, an enterprise agent blueprint positioned to cut deployment costs and speed enterprise AI adoption.
SpaceXAI has launched Grok 4.5, positioning the model as a cheaper, faster rival for coding and knowledge work ahead of a new OpenAI release.
Prime Intellect raised a $130 million Series A to sell infrastructure for enterprise-built AI agents, betting buyers want more control than frontier labs offer.
