Moonshot AI (Kimi)

Kimi K2 Thinking Turbo

Type
language model
Context
262K tokens
Max output
262K
Released
5 September 2025
Knowledge cutoff
January 2025
API string
kimi-k2-thinking-turbo
Weights
open

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
moonshotai-cn $1.15 $8 31 Jul 2026
Moonshot AI $1.15 $8 31 Jul 2026
ZenMux $1.15 $8 31 Jul 2026
302.AI $1.265 $9.119 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 1,453.31 1,448–1,459
Arena Score, expert questions 1,438.16 1,429–1,448
Arena Score, mathematics 1,429.15 1,419–1,439
Arena Score in English 1,428.12 1,424–1,432
Arena Score, hard prompts 1,427.43 1,423–1,431
Arena Score in French 1,424.58 1,408–1,442
Arena Score, long queries 1,414.57 1,409–1,420
Arena Score, overall 1,414.01 1,411–1,417
Arena Score, multi-turn dialogue 1,413.95 1,408–1,420
Arena Score in Spanish 1,407.87 1,392–1,423
Arena Score, instruction following 1,401.9 1,397–1,407
Arena Score in Russian 1,390.37 1,383–1,398
Arena Score, creative writing 1,374.88 1,368–1,382
Arena Score, web development 1,322 1,315–1,329
Mock AIME 2024–2025 — olympiad problems 83.04 % · with a tuned harness Epoch evaluations
GPQA Diamond — graduate-level questions 78.96 % · with a tuned harness Epoch evaluations
SimpleQA Verified — factual accuracy 31.6 % Epoch evaluations
Chess puzzles 15.83 % Epoch evaluations