Moonshot AI (Kimi) · Flagship

Kimi K3 API pricing and specs

Moonshot's open-weight model with always-on reasoning, native vision, dynamically loaded tools and JSON Schema output.

Input

⁨$3⁩

Cached input

⁨$0.30⁩

Output

⁨$15⁩

US dollars per million tokens, standard tier, short-prompt rate

Default reasoning effort is max, which is expensive in output tokens; set it lower for routine steps. Moonshot advises against its built-in web search while it is being updated.

Spec sheet

API model string: kimi-k3
Provider: Moonshot AI (Kimi)
Context: 1.05M tokens
Max output: —
Released: July 16, 2026
Batch API: No batch discount listed
Documented agent features: Tool callingStructured outputAdjustable reasoningImage input

What a task costs

Four reference workloads priced at this model's list rates, with caching applied after the first call. Your numbers will differ; the calculator lets you change every assumption.

Workload	Per task	Per 1,000 tasks
Support bot 4 calls, short tool lookups	⁨$0.084⁩	⁨$84⁩
Coding agent 30 calls, files and test output	⁨$2.97⁩	⁨$2,966⁩
Browser agent 25 calls, screenshots and page text	⁨$2.09⁩	⁨$2,095⁩
Research agent 15 calls, long search results	⁨$1.98⁩	⁨$1,985⁩