🇨🇳 China💻 CodeReleased April 2026
Kimi K2.6
Moonshot AI
Open-sourceFreeSelf-host
Verdict
A solid pick for self-hosting and full data control.
Overview
Kimi K2.6 is Moonshot AI's flagship open-weight model: 1 trillion total parameters, 32B active per token (MoE, 384 experts), 256K context. Built for long-horizon agentic coding, orchestration of hundreds of sub-agents and full-stack generation. Multimodal (text, image, video), Modified MIT license, and very low pricing versus closed models. Stronger on agentic tasks than on pure reasoning.
Skill profile
Verified benchmarks
GPQA90.5%
What do these scores mean?
GPQA90.5%world-class
PhD-level science questions (physics, chemistry, biology), with no tool access.
Strengths
- Has a free tier
- Open-source and self-hostable
- Competitive input pricing
Limitations
- Average speed
- No native GDPR guarantee
- Light safety filters
Who is it for
A good fit if…
- you build with a coding agent
- you want to control cost or self-host
Skip it if…
- you handle sensitive EU data
- you need very fast responses
- you want strict guardrails
Ideal use cases
- Long-horizon agentic coding and multi-agent orchestration
- Full-stack generation (UI + backend)
- Chinese and multilingual workloads
Access & availability
Paid API per tokenFree tier availableSelf-hostable (open weights)
Key specifications
Context
262K
Input price
$0.95 $/M
Output price
$4.00 $/M
Speed
—
Price verified Jul 2, 2026·Official source ↗Benchmarks not audited
Estimate your monthly cost
Per-token API pricing$57/ month with Kimi K2.6
For the same usage
- DeepSeek Coder V3$6.19-89%
- Codestral$15-73%
Indicative estimate based on standard API rates (excluding caching, batch and volume discounts). Always check the official pricing before committing.
Privacy
Not GDPR-compliant
Advanced data · for expertsArchitecture, modalities, detailed cost, full benchmarks▾
Advanced data · for experts
Architecture, modalities, detailed cost, full benchmarks
Architecture
MoE (384 experts, 32B actifs)
Size
1T total / 32B actifs (MoE)
Cutoff date
—
Inputs
text, image, video
Outputs
text
License
mit
Hosting
—
Hallucination score
—
Estimated cost (API)
Typical exchange (~3k in / 1k out)
≈ $0.0069
1M in + 1M out
≈ $4.95
All benchmarks
| Arena Elo | — |
| MMLU | — |
| GPQA | 90.5% |
| HumanEval | — |
| SWE-Bench | — |
| MATH | — |