China AI Hub China AI Hub

Models / Reasoning Chinese AI models

Reasoning Chinese AI models

Reasoning models interleave an extended chain-of-thought before answering, improving performance on math, code and hard knowledge tasks at the cost of latency and output tokens. This hub lists every model that records a reasoning capability flag.

22
models match this dimension
Counted automatically from the models database: capabilities.reasoning === true

Reasoning is near-universal across the 2026 lineup. The real differentiators are whether it can be disabled, its effort levels (low/high/max) and its price — not whether it exists.

Editorial position

Our read: reasoning models are not more expensive per token — they are more expensive per answer, sometimes by an order of magnitude, because they emit hundreds of hidden tokens before the first visible one. Buy them where a wrong answer costs more than the tokens: code review, financial extraction, multi-step planning. Not for classification, where a flash tier is far cheaper and equally accurate.

Pattern in the data

Distribution of this dimension across matching models
Models matching22
Open weights15 of 22
Providers represented7
Context window range200K – 1024K
Input price range /1M$0.15 – $26
Licence familiesMIT, proprietary, Apache-2.0, Modified, Kimi, Custom, MiniMax, Qwen3.8-Max

Reasoning models cluster at the top of the price range, and their cost is better measured per answer than per token: hidden reasoning tokens are billed as output. In the data, every reasoning model costs several times the flash tier of the same family per visible token of answer.

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-M2.7

minimax · active

  • Context: 204,800 tokens
  • Open weight · Custom NON-COMMERCIAL license (MIT-style terms for non-commercial use only; any commercial use requires prior written authorization from MiniMax at [email protected]; attribution 'Built with MiniMax M2.7' required)

MiniMax-M2.7-Highspeed

minimax · active

  • Context: 204,800 tokens
  • Open weight · Custom NON-COMMERCIAL license (MIT-style terms for non-commercial use only; any commercial use requires prior written authorization from MiniMax at [email protected]; attribution 'Built with MiniMax M2.7' required)

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])

Qwen3.8-2.4T-A95B

alibaba-cloud · active

  • Context: 262,144 tokens
  • Open weight · Qwen3.8-Max License (custom MIT-style: unrestricted use/copy/modify/sell, but products with >100M MAU or >US$20M/month revenue must display the model name; Model-as-a-Service or AI Work Assistant businesses with >US$50M/12-month revenue need a separate license from Qwen)

Decision guides

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.