All tags
Model: "devstral-small-2"
MCP -> Agentic AI Foundation, Mistral Devstral 2
devstral-2 devstral-small-2 sonnet-4.3 deepseek-v3.2 qwen3-vl openai anthropic block mistral-ai alibaba linux-foundation deepseek agentic-ai coding-models reinforcement-learning model-performance model-optimization open-weights cli-tools multi-file-code-automation data-decontamination moe reward-models rl-stability guillaumelample b_roziere qtnx_ charliermarsh omarsar0 eliebakouch justinwaugh cwolferesearch pan
OpenAI Engineering sees a significant collaborative milestone with the launch of the Agentic AI Foundation under the Linux Foundation, uniting projects from Anthropic, OpenAI, and Block. Mistral released Devstral 2, a coding model with 123B parameters and open weights, offering a cost-effective alternative to Sonnet 4.3 and competitive performance against DeepSeek v3.2. The new Mistral Vibe CLI supports agentic coding workflows with rapid ecosystem integration. Alibaba introduced Soft Adaptive Policy Optimization (SAPO) for reinforcement learning tuning, improving stability and performance in Qwen3-VL across multiple tasks. Research highlights include the importance of data decontamination in RL and ongoing discussions on MoE RL stability and reward hacking mitigation.