Our take
Moonshot AI sells four offers, all of its own Kimi models, and states both a no-training policy and a zero-retention option. Nothing here measures how well those models actually perform, so the decision rests on terms and price rather than on evidence.
Use it for work whose prompts must not be kept or reused, since both a no-training policy and a zero-retention option are stated, and for cost-sensitive coding or chat on the cheaper Kimi offers. The same model is sold at two rates, so check which offer you are picking before you commit. Skip it if you need an independently verified security attestation, a confirmed European endpoint, or published quality evidence before you commit.
- Prompt handling is stated rather than left open: no training on prompts, plus a zero-retention option, which is what you need when the work cannot go to a host that keeps it.
- Kimi K2.7 Code is sold at two different rates, so choosing the lower-priced offer is a real decision rather than a rounding difference.
- No benchmark score is supplied for any of its models, so nothing here tells you how well they code, reason or follow instructions; a trial on work you can judge yourself is the only way to find out.
- An independent security attestation and a European endpoint are both unverified in our data, and the headquarters country is not disclosed, so regulated or data-residency-bound work cannot be cleared from these facts alone.
Point your tools here
We have not recorded what a router needs for Moonshot AI yet — no base URL and no docs link on file. That is our gap, not a sign Moonshot AI has no API; its own documentation is the place to look until we close it.
Privacy & data handling
These answers cover requests sent to Moonshot AI through OpenRouter, as OpenRouter records them, last read 3 hours ago. For Moonshot AI's own API we hold no answer: check its terms. SOC 2 and the data processing agreement below are hand-checked, not part of that sync.
Prompt retention: none.