Kimi K2.7 Code

ModelOpen sourceActive

by Moonshot AI · family “kimi-k2-code”

MoonshotAI: Kimi K2.7 Code is a coding-focused model in Moonshot AI's Kimi K2 family, built to complete end-to-end programming tasks reliably over long contexts. It uses a native multimodal mixture-of-experts... (description from the OpenRouter listing)

in: textin: imageout: textReasoningTool useStructured output

Epoch Capabilities Index

150.0 #48 of 268

90% CI 148.1 – 151.7

Listing ↗

Released

12 Jun 2026

Context window

262K tokens

Max output

236K tokens

Input $ / 1M tokens

$0.66

Output $ / 1M tokens

$3.30

Cached input $ / 1M

$0.18

Each benchmark shown separately with its own source

Benchmark results (12)

All benchmarks

Benchmark	Domain	Score	vs best recorded	Setting	Run	Source
Surface Evolver Bench	science	48.8%	51%	—	—	External ↗
FrontierMath-Tiers-1-3-v2-Private	math	54.0% ±3.0	58%	—	13 Jun 2026	Eval log ↗
FrontierMath-Tier-4-v2-Private	math	12.2% ±5.2	12%	—	13 Jun 2026	Eval log ↗
FrontierCode	coding	30.1%	56%	—	—	External ↗
DeepSWE	coding	30.5%	41%	—	—	External ↗
APEX-Agents	agents	37.6%	55%	—	—	External ↗
Chess Puzzles	games	21.0% ±4.1	29%	—	7 Aug 2026	Eval log ↗
SimpleQA Verified	knowledge	36.5% ±1.5	48%	—	27 Aug 2026	Eval log ↗
WeirdML	coding	54.1% ±0.0	58%	—	—	External ↗
OTIS Mock AIME 2024-2025	math	95.6% ±3.1	96%	—	7 Aug 2026	Eval log ↗
SimpleBench	reasoning	57.9%	71%	—	—	External ↗
GPQA diamond	science	87.9% ±2.3	92%	—	7 Aug 2026	Eval log ↗

Source: Epoch AI Benchmarking Hub (CC BY 4.0). “External” rows are leaderboard results Epoch collects from third parties. Best reported setting per benchmark is shown.

API list price over time

Price history

$ per 1M tokens

4 observations (OpenRouter listing + Internet Archive snapshots). History accumulates with every ingest run; a single point means no change has been observed yet.