China AI Hub China AI Hub

Models / Multimodal Chinese AI models

Multimodal Chinese AI models

Multimodal models accept non-text input — images, and in some cases video or screen control. This hub lists every model that records a vision-input flag, with the rarer video and computer-use flags shown where documented.

12
models match this dimension
Counted automatically from the models database: capabilities.vision === true

Vision has migrated down the price stack: the cheapest tiers are now among the most capable, and two flagships (DeepSeek-V4-Pro, GLM-5.3) are text-only while their $0.15 flash siblings carry vision.

Editorial position

Our read: vision claims are broadly true and video claims are mostly untested. Image understanding is reliable enough to build on; video is where demos outrun published evaluation, and computer use is where the gap between demo and product is widest. Treat any multimodal claim without a named benchmark and version as marketing.

Pattern in the data

Distribution of this dimension across matching models
Models matching12
Open weights7 of 12
Providers represented6
Context window range256K – 1024K
Input price range /1M$0.15 – $26
Licence familiesMIT, proprietary, Apache-2.0, Modified, Kimi, MiniMax

The data separates vision from video and from computer use, and the three do not move together: many models accept images, fewer handle video, and fewer still support screen control. Buying 'multimodal' on the strength of an image demo is the most common over-specification in this dataset.

Matching models

GLM-5.3-Flash

zhipu-ai · active

  • Context: 1,048,576 tokens
  • Open weight · Apache-2.0 (per GitHub repo metadata; README has no separate weights-license section - verify per-model HF cards before reuse)

Kimi K3

moonshot-ai · active

  • Context: 1,048,576 tokens
  • Open weight · Kimi K3 License (permissive MIT-style, but Model-as-a-Service operators with >$20M aggregate revenue over any 12 months must sign a separate agreement; products with >100M MAU or >$20M monthly revenue must display 'Kimi K3' in the UI)

MiniMax-M3

minimax · active

  • Context: 1,048,576 tokens
  • Open weight · MiniMax Community License (custom): free for non-commercial use; commercial use requires prominent attribution 'Built with MiniMax M3' plus written authorization from MiniMax if yearly revenue exceeds US$20M (otherwise a one-time notice to [email protected])

Decision guides

Original research

Technology

This page is generated from the models database; the member list and count are derived from each model's frontmatter fields, not written by hand. See the full model database and the methodology.