Moonshot AI (Kimi)

Kimi K2.7 Code

Type
language model
Context
262K tokens
Max output
262K
Released
12 June 2026
API string
kimi-k2.7-code
Weights
open

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
routing-run $0.275 $1.1 31 Jul 2026
crof $0.55 $2.25 31 Jul 2026
Weights & Biases $0.71 $3.5 31 Jul 2026
ambient $0.73 $3.5 31 Jul 2026
OpenRouter $0.73 $3.5 31 Jul 2026
DeepInfra $0.74 $3.5 31 Jul 2026
aiand $0.75 $3.5 31 Jul 2026
inceptron $0.75 $3.5 31 Jul 2026
Venice AI $0.75 $3.5 31 Jul 2026
greenpt $0.941 $4.389 31 Jul 2026
opencode-go $0.95 $4 31 Jul 2026
cloudflare-workers-ai $0.95 $4 31 Jul 2026
cline-pass $0.95 $4 31 Jul 2026
moonshotai-cn $0.95 $4 31 Jul 2026
neuralwatt $0.95 $4 31 Jul 2026
synthetic $0.95 $4 31 Jul 2026
hyper $0.95 $4 31 Jul 2026
novita-ai $0.95 $4 31 Jul 2026
perplexity-agent $0.95 $4 31 Jul 2026
hpc-ai $0.95 $4 31 Jul 2026
AIHubMix $0.95 $4 31 Jul 2026
Baseten $0.95 $4 31 Jul 2026
Cloudflare Workers AI $0.95 $4 31 Jul 2026
Databricks $0.95 $4 31 Jul 2026
EmpirioLabs AI $0.95 $4 31 Jul 2026
Fireworks AI $0.95 $4 1 Aug 2026
GitHub Copilot $0.95 $4 31 Jul 2026
Hugging Face $0.95 $4 31 Jul 2026
LLM Gateway $0.95 $4 31 Jul 2026
Merge Gateway $0.95 $4 31 Jul 2026
Moonshot AI $0.95 $4 31 Jul 2026
Nebius Token Factory $0.95 $4 31 Jul 2026
OpenCode Zen $0.95 $4 31 Jul 2026
Pioneer $0.95 $4 31 Jul 2026
Together AI $0.95 $4 31 Jul 2026
Vercel $0.95 $4 31 Jul 2026
ZenMux $0.95 $4 31 Jul 2026
CrossModel $1 $4.16 31 Jul 2026
tensorx $1.25 $4.5 2 Aug 2026
Cortecs $1.28 $4.63 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, web development 1,472.57 1,463–1,482
Mock AIME 2024–2025 — olympiad problems 96.39 % · with a tuned harness Epoch evaluations
GPQA Diamond — graduate-level questions 86.03 % · with a tuned harness Epoch evaluations
WeirdML — unusual machine learning tasks 54.12 % https://htihle.github.io/weirdml.html
FrontierMath, levels 1–3 54.04 % Epoch evaluations
SimpleBench — trick questions 49.48 % SimpleBench Leaderboard
SimpleQA Verified — factual accuracy 39.2 % Epoch evaluations
FrontierCode — patches fit to be merged into a project 30.1 % https://cognition.com/frontiercode
APEX-Agents 27.6 %
Chess puzzles 16.88 % Epoch evaluations
FrontierMath, level 4 — research-grade problems 12.2 % Epoch evaluations
CritPt — physics problems 10 %
Arena Score, agent tasks 0 0–0