Kimi token counter
Paste text to estimate how many tokens it uses and what sending it costs on each Kimi model. Everything runs in your browser — nothing you type is transmitted or stored anywhere.
Last updated
- Estimated tokens
- 41
- Characters
- 164
- Words
- 25
- Chars / token
- 4
Estimated at 4 characters per token. Expect roughly 10–15% error on English prose, and more on code or non-English text.
Context window used
- Kimi K3<0.01%
- Kimi K2.7 Code<0.01%
- Kimi K2.7 Code HighSpeed<0.01%
- Kimi K2.6<0.01%
- Kimi K2.5<0.01%
Cost
What this text costs to send
Input cost only, at the cache-miss rate — the price you pay the first time the model sees this text. Output is billed separately.
| Model | Input cost | Cost if cached |
|---|---|---|
| Kimi K3 | — | — |
| Kimi K2.7 Code | — | — |
| Kimi K2.7 Code HighSpeed | — | — |
| Kimi K2.6 | — | — |
| Kimi K2.5 | — | — |
How many characters is one Kimi token?
Moonshot documents roughly 3 to 4 characters per token for typical English text. This tool uses 4, the conservative end, so it tends to under-count slightly rather than surprise you with a larger bill.
How accurate is this token counter?
It is an estimate, not an exact count. Moonshot has not published its tokenizer, so no browser-based tool can be exact. For English prose expect roughly 10 to 15 percent error. For code, JSON or non-English text the error is larger, because those tokenize less efficiently.
How do I get an exact Kimi token count?
Moonshot exposes a token estimation endpoint on the Kimi API platform that returns the exact count for a given set of messages and model. Use that when you need precision — for example when you are close to a context window limit.
Is my text sent anywhere?
No. The counting runs entirely in your browser using JavaScript on this page. Nothing you paste is transmitted, stored or logged. You can verify this by opening your browser network tab while typing.
Do input and output tokens count toward the same limit?
Yes. A model’s context window covers input and output combined. Kimi K3’s 1,048,576-token window means your prompt plus the response must fit within that total, so a very long prompt reduces the room left for the answer.