Models

October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding

NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use

Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.

LightOnOCR-3-0.8B is the lightest model of LightOn's LightOnOCR-3 family, with the same layout and visual understanding features as its larger siblings.

First MoE MiniCPM-V model for single-image, multi-image and video understanding

Compact MiniCPM-V model for single-image, multi-image and video understanding

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

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

MiMo pro model for strong multimodal reasoning and agent execution

StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window

DeepSeek V4.1 Flash model for reasoning and agentic coding

Native multimodal hybrid MoE model for image and video understanding, reasoning, and visual agents

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

Native multimodal GLM model for efficient coding and long-horizon agent tasks

Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work

Mixture-of-experts coding-reasoning model for agentic software tasks, tool use, and image understanding

Dense 27B vision-language model for coding, agent tasks, and image and video understanding

Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio

Qwen3.6-based Mixture-of-LoRA model with specialists for chat, personal agents, coding, and generative UI

Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work

Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio

Open coding-reasoning model for repository tasks and self-improving agents

Large coding-reasoning model for agentic software tasks and RL search

Open coding-reasoning model for repository tasks and self-improving agents

Large coding-reasoning model for agentic software tasks and RL search

Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking

Lower-latency Kimi Code variant for interactive edits and coding-agent loops

Compact Gemma 4 instruction model for open, self-hosted chat and reasoning

Gemma 4 based discrete text diffusion model for fast chat, reasoning, and tool use

Open agentic MoE model (397B total, 17B active) for coding, tool use, and research workflows

MiniMax multimodal model for long-context coding, perception, and agent planning

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

Balanced Mistral model for enterprise assistants, multilingual work, and tools

Balanced Mistral model for enterprise assistants, multilingual work, and tools

Open Nemotron omni model combining reasoning with text, vision, and audio

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

Qwen vision-language model for visual reasoning, documents, and agent tasks

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

Open multimodal Qwen MoE for local agents that need vision, audio, and code

Open Gemma instruction model for efficient chat and self-hosted deployments

Largest Gemma 4 instruction model for open, self-hosted chat and reasoning

Efficient Mistral model for fast chat, extraction, and production assistants

Fast Mistral production model for chat, extraction, and cost-sensitive agents

Compact Qwen3.5 multimodal model for fast chat, reasoning, and tool use

Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use

Smallest Qwen3.5 multimodal model for on-device chat, reasoning, and tool use

Pre-trained Qwen3.5 base checkpoint for fine-tuning and research, without instruction tuning

Pre-trained Qwen3.5 base checkpoint for fine-tuning and research, without instruction tuning

Qwen vision-language model for visual reasoning, documents, and agent tasks

Qwen vision-language model for visual reasoning, documents, and agent tasks