Kimi K2.6

ModelOpen sourceActive

by Moonshot AI · family “kimi-k”

Kimi K2.6 is Moonshot AI's next-generation multimodal model, designed for long-horizon coding, coding-driven UI/UX generation, and multi-agent orchestration. It handles complex end-to-end coding tasks across Python, Rust, and Go, and... (description from the OpenRouter listing)

in: textin: imageout: textReasoningTool useStructured output

Epoch Capabilities Index

151.0 #44 of 268

90% CI 149.2 – 152.8

Listing ↗

Released

20 Apr 2026

Context window

262K tokens

Max output

236K tokens

Input $ / 1M tokens

$0.65

Output $ / 1M tokens

$3.41

Cached input $ / 1M

$0.15

Each benchmark shown separately with its own source

Benchmark results (18)

All benchmarks

Benchmark	Domain	Score	vs best recorded	Setting	Run	Source
Furniture Assembly	multimodal	21.7% ±5.2	26%	—	11 Sep 2026	Epoch ↗
LMCA	agents	37.3%	57%	—	—	External ↗
DTBench	reasoning	90.9%	92%	—	—	External ↗
Mystery Game Puzzles	games	18.0% ±3.9	21%	—	17 Jul 2026	Epoch ↗
EBR-bench	reasoning	2.4%	3%	—	25 Jun 2026	Epoch ↗
OSWorld 2.0	agents	4.6%	15%	—	—	External ↗
FrontierMath-Tiers-1-3-v2-Private	math	57.2% ±2.9	61%	—	10 Jun 2026	Eval log ↗
FrontierMath-Tier-4-v2-Private	math	25.6% ±7.1	26%	—	10 Jun 2026	Eval log ↗
ExploitBench	coding	18.4%	25%	—	—	External ↗
ProofBench	math	16.0%	16%	—	—	External ↗
Chess Puzzles	games	26.0% ±4.4	36%	—	7 May 2026	Eval log ↗
SimpleQA Verified	knowledge	34.9% ±1.5	46%	—	10 Aug 2026	Epoch ↗
FrontierMath-Tier-4-2025-07-01-Privatesuperseded	math	14.6% ±5.1	30%	—	8 May 2026	Epoch ↗
FrontierMath-2025-02-28-Privatesuperseded	math	39.0% ±2.9	74%	—	7 May 2026	Epoch ↗
WeirdML	coding	55.9% ±0.0	60%	—	—	External ↗
OTIS Mock AIME 2024-2025	math	96.1% ±2.4	96%	—	2 May 2026	Eval log ↗
SWE-Bench verified	coding	76.7% ±1.9	92%	—	8 May 2026	Eval log ↗
GPQA diamond	science	90.8% ±1.7	95%	—	1 May 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

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