DeepSeek-V4-Flash-Vision-Exp is an experimental vision-enabled variant of DeepSeek V4 Flash. It accepts text and image inputs and produces text, supporting image description, screenshot text recognition, chart analysis, visual reasoning, coding, and agent workflows.
Flagship GLM model for long-horizon coding, agents, and complex project delivery
Wan3.0-Video is an all-in-one video generation model from Alibaba. It accepts text, image, video, and audio inputs for reference-based generation, editing, replication, and motion-driven workflows, and generates videos up to 30 seconds. It can also understand files, webpages, and complex images.
2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows
Video model for prompt-guided generation, editing, and motion workflows
Qwen3.7-Flash is an efficient native vision-language model from Alibaba Cloud for text, image, and video understanding, with enhanced multimodal reasoning, agent execution, and coding capabilities.
Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work
hy-3d-3.1 is a model from Tencent for text- and image-guided 3D asset generation.
Qwen role-playing model optimized for persona fidelity, instruction following, proactive topic progression, and empathetic dialogue.
Tencent Hy reasoning model for coding, instruction following, and agent tasks
doubao-seed-2-1-pro-260628 is a model from ByteDance for text generation, reasoning, and assistant workflows.
doubao-seed-2-1-turbo-260628 is a model from ByteDance for text generation, reasoning, and assistant workflows.
doubao-seedance-2-5-260628 is a model from ByteDance for prompt- and image-guided video generation.
doubao-seedream-5-0-pro-260628 is a model from ByteDance for image generation.
doubao-seedance-2-0-mini-260615 is a model from ByteDance for prompt- and image-guided video generation.
Open flagship GLM for long-horizon coding agents and million-token context work
Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking
Multimodal Qwen workhorse for long-context agents, visual inputs, and coding
MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding, and tool use. It is built on MiniMax Sparse Attention (MSA), which replaces full attention with KV-block selection to cut per-token compute at long context — roughly 1/20 the cost of the previous generation at 1M tokens, with substantially faster prefill and decode while retaining quality across most tasks.
Multimodal Qwen workhorse for long-context agents, visual inputs, and coding (site alias/snapshot: qwen3.7-plus-2026-05-26).
