Browse Chinese LLMs by capability

What do you need the model to do? Chinese AI has specialists for every bucket — video, reasoning, long-context document Q&A, code agents. Pick a capability to see which models lead it and who hosts them.

Text

Chat106 models
General conversation, instruction following, multilingual
Reasoning49 models
Step-by-step thinking, chain-of-thought, tool-use
Code43 models
Code generation, completion, agentic tool use
Math4 models
Mathematical reasoning and proofs
Long context15 models
128K-1M+ context windows for document Q&A

Image & video

Multimodal (vision)62 models
Vision-language models that accept image input
Text-to-image14 models
Generate images from text prompts
Image editing8 models
Edit reference images or decompose designs into transparent layers
Video generation16 models
Generate video from text, images and other references

Retrieval

Embedding5 models
Text embeddings for semantic search and RAG
Reranking4 models
Score and reorder search candidates for RAG

Audio

Text-to-speech5 models
Generate speech from text with hosted or local models
Speech-to-text2 models
Transcribe audio with speech-recognition models