Kimi K3 Price Hike: Output Jumps 11% to $15 per Million Tokens
The provider of Kimi K3 has raised output token pricing by 11% to $15.00 per million, while input drops slightly to $0.70. Here is how it impacts generation costs.
👁 126 reads
The provider behind Kimi K3 has adjusted its pricing. Input tokens now cost $0.70 per million (down from $0.73), while output tokens have risen to $15.00 per million (up from $13.50). That marks an 11% increase on the item that carries the most weight in standard usage bills.
| Rate per million tokens | Previous | Current | Change |
|---|---|---|---|
| Input | $0.73 | $0.70 | -4% |
| Output | $13.50 | $15.00 | +11% |
Output tokens carry the heaviest weight on overall invoices because model responses are usually much longer than the input prompt. That is why when evaluating API rates, looking at the output figure first provides a clearer sense of real-world costs than focusing solely on input pricing.
In practical terms, generating a 1,500-word article with this model costs around $0.03. For comparison, the cheapest models tracked on the list are currently Qwen 3.7 Flash ($0.13 per million output tokens) and DeepSeek V4 Flash ($0.64).
A complete overview of rates across tracked models, broken down by practical tasks, is updated daily in the AI cost calculator. Pricing is sourced from the public OpenRouter API and may vary from promotional deals offered directly by the provider.
Google · Your sourcesQuantslant higher in your Google - one clickAdd to preferred sources →
More on this from QUANTSLANT
magazine·Thélyson Orélien and the AI Detector: I Tested the Prize-Winning Novel. Same Paragraph Scored 100%, 85% and 0% AIThe Thélyson Orélien affair explained, plus my own test of the novel's excerpts in Pangram, AUTHORRA and ZeroGPT, with Zola and an AI text as controls.