China AI Hub China AI Hub

Models / Coding Chinese AI models

Coding Chinese AI models

Coding models are documented for code generation and agentic software work — the arena where China's frontier race is most intense. This hub lists every model that records a coding capability flag.

14
models match this dimension
Counted automatically from the models database: capabilities.coding === true

Coding evidence is unevenly published: SWE-bench and Terminal-Bench versions differ across vendors, so the coding flag plus each model's documented benchmark scores is the honest signal set — not a single ranking.

Editorial position

Our read: coding is the least comparable category in the market. Vendors mix single-shot and agentic benchmarks, report pass@1 and pass@k interchangeably, and rarely state the harness. A model leading a public leaderboard can lose on your repository, because the variable that matters — how well it recovers from a failed test — is almost never published.

Pattern in the data

Distribution of this dimension across matching models
Models matching14
Open weights11 of 14
Providers represented6
Context window range256K – 1024K
Input price range /1M$0.15 – $26
Licence familiesMIT, Apache-2.0, Kimi, MiniMax, proprietary

Coding capability is the most evenly distributed flag in the dataset — almost every frontier model claims it — which is precisely why the flag has little selection value. The differentiators in the data are context window, price and licence, not the coding label.

Matching models

GLM-5.2

zhipu-ai · deprecated

  • Context: 1,048,576 tokens
  • Open weight · MIT (pure open, no regional limits per official model card)

GLM-5.3

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)

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

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.