Moonshot AI (Kimi)

Kimi K2.5 Thinking TEE

Type
language model
Context
256K tokens
Max output
66K
Released
29 January 2026
API string
TEE/kimi-k2.5-thinking

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
NanoGPT $0.3 $1.9 31 Jul 2026

Only the base tier and only per-token prices are shown. Batch, discounted and cached rates, as well as prices per image or per second of video, do not go into this table: they cannot stand in the same column as a price per million tokens.

The date in the last column is the day this price first entered our collection. The price may well be older: before that day we simply were not recording it. It has not changed since — otherwise a new row with a new date would stand in its place.

Measurement results

Task set Result Run conditions Measured by
Arena Score, programming 1,473.6 1,468–1,479 — —
Arena Score, mathematics 1,470.79 1,461–1,481 — —
Arena Score, expert questions 1,466.31 1,458–1,475 — —
Arena Score, hard prompts 1,454.18 1,450–1,458 — —
Arena Score in English 1,453.91 1,449–1,458 — —
Arena Score in Spanish 1,453.89 1,439–1,469 — —
Arena Score in French 1,453.7 1,439–1,468 — —
Arena Score, overall 1,445.19 1,442–1,449 — —
Arena Score, long queries 1,445.11 1,440–1,450 — —
Arena Score, multi-turn dialogue 1,443.94 1,438–1,450 — —
Arena Score in Russian 1,436.38 1,429–1,444 — —
Arena Score, web development 1,435.77 1,430–1,441 — —
Arena Score, instruction following 1,431.01 1,426–1,436 — —
Arena Score, creative writing 1,423.84 1,417–1,431 — —
Arena Score, understanding diagrams 1,280.8 1,272–1,289 — —
Arena Score, text recognition in images 1,273.54 1,268–1,279 — —
Arena Score, working with images 1,265.7 1,260–1,272 — —