# Which Chinese AI > Dated, sourced guide to Chinese AI prices, access, and privacy for readers outside China. Figures in articles are rendered from JSON snapshots. There are no affiliate links. ## Primary pages - https://whichchineseai.com/prices/ : live OpenRouter price table - https://whichchineseai.com/calculator/ : illustrative cost per task - https://whichchineseai.com/methodology/ : blended price, comparators, benchmark files - https://whichchineseai.com/disclosure/ : no affiliate links - https://whichchineseai.com/privacy/ : site privacy - https://whichchineseai.com/about/ : anonymous brand - https://whichchineseai.com/corrections/ : corrections log ## Articles - https://whichchineseai.com/guides/chinese-models-outside-china/ : OpenRouter's endpoint list shows US-named hosts for GLM-5.3 and Kimi K3. The cheap host is not automatically the non-China host. - https://whichchineseai.com/guides/chinese-ai-api-prices/ : OpenRouter prices for Chinese models, which flash tiers are actually cheap, and where the list price is not the lab's rate. - https://whichchineseai.com/guides/coding-plans-compared/ : GLM publishes credits and a standard price on its own docs. DeepSeek's API page is pay-as-you-go. Other plan prices were not re-verified. - https://whichchineseai.com/guides/deepseek-api-key/ : The docs bill a topped-up balance. Off-peak is half. We did not confirm which cards work from abroad. - https://whichchineseai.com/guides/deepseek-v4-flash-openrouter/ : OpenRouter's V4.1 Flash card matches DeepSeek's peak rate. Off-peak direct is half. Here is the setup and the catch. - https://whichchineseai.com/compare/deepseek-vs-claude/ : V4 Flash is cheap on DeepSeek's own page. The public SWE-bench file still does not list V4. Here is the gap that is real. - https://whichchineseai.com/apps/doubao/ : Doubao is the consumer app. Seed ids on OpenRouter are a different product. We did not install the app or invent an English feature list. - https://whichchineseai.com/apps/free-chinese-ai-apps/ : DeepSeek's chat app is free on the official stores. Qwen Chat is a website. Video apps need a credit check we have not stored. - https://whichchineseai.com/guides/glm-coding-plan/ : Z.AI's coding plan starts at $18 on the migration notice, with 5-hour and weekly credit caps. What the docs actually say. - https://whichchineseai.com/apps/hailuo/ : Hailuo is MiniMax video, not the M3 token price. We did not store a credit grid, so we do not invent one. - https://whichchineseai.com/guides/is-deepseek-free/ : The chat app is free. The API is not. Here is what the store listings and the pricing page actually say. - https://whichchineseai.com/guides/is-chinese-ai-safe/ : What Czech, Taiwanese, and Reuters reporting actually said about DeepSeek, and the three ways to send a prompt. - https://whichchineseai.com/compare/kimi-vs-claude/ : Kimi K3 is close on one public preference board and expensive on output. Claude still leads the rows we can actually cite. - https://whichchineseai.com/guides/is-kimi-k3-open-source/ : Moonshot released Kimi K3 weights under the Kimi K3 License, with revenue thresholds for API businesses and large products. - https://whichchineseai.com/guides/kimi-k3/ : Moonshot lists Kimi K3 at $3 input and $15 output. OpenRouter's headline price is a different host. Weights are public, with a licence catch. - https://whichchineseai.com/apps/kling/ : Kling is Kuaishou's video tool. We did not capture a stable credit price on 16 Oct 2026, so this page refuses to invent one. - https://whichchineseai.com/guides/mimo-stepfun-hunyuan/ : MiMo V2.6 Pro and Hunyuan HY3 are inexpensive on OpenRouter. We do not have lab invoices or SWE-bench rows for them. - https://whichchineseai.com/guides/minimax-m3/ : M3 is cheap on the OpenRouter blend. That chart is not a Hailuo video price and not a SWE-bench win we could find for M3. - https://whichchineseai.com/guides/openrouter-vs-direct/ : DeepSeek off-peak undercuts OpenRouter. GLM-5.3 and Kimi K3 list prices are cheap hosts, not the labs. - https://whichchineseai.com/guides/paying-from-abroad/ : USD prices exist for Z.AI, DeepSeek, and Kimi. Which foreign cards clear was not tested. No reseller links. - https://whichchineseai.com/guides/qwen-api-free-tier/ : Model Studio and the China console are different accounts. We did not confirm a free-token number, so this page does not print one. - https://whichchineseai.com/apps/qwen-chat/ : Use the official Qwen chat site or store listing. A modified APK is not a free tier. The API is a separate product. - https://whichchineseai.com/compare/which-chinese-model/ : A task matrix using OpenRouter prices and the public boards we stored. No single winner, and two prices you should not trust alone. - https://whichchineseai.com/guides/run-qwen-locally/ : Smaller Qwen ids have a public API price you can compare with a GPU you already own. We do not invent a tokens-per-second figure. - https://whichchineseai.com/digest/2026-w42/ : Week of 16 Oct 2026. The snapshot, the GLM and Kimi endpoint check, and the benchmark files we stored. - https://whichchineseai.com/guides/claude-code-chinese-models/ : Anthropic-compatible base URLs for Z.AI and DeepSeek, plus what we would not paste in for Kimi or Qwen. - https://whichchineseai.com/guides/coding-session-cost/ : An illustrative 3 million token session, priced from the snapshot. Cache-write fees are excluded and the mix is not your repo. - https://whichchineseai.com/guides/z-ai-api/ : GLM-5.3 Flash matches the docs at $0.15 / $0.50. The flagship is $1.40 / $4.40, not the OpenRouter headline.