Models

Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding

Tencent Hy reasoning model for coding, instruction following, and agent tasks

Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution

Newer StepFun flash model for faster agents, coding, and multimodal prompts

Open MiMo model for multimodal coding agents and long-context automation

Open MoE flagship with million-token context for coding and long agent runs

Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work

Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.

Multimodal Kimi workhorse for agent loops, coding tasks, and visual context