China AI Hub
Models / Kimi K2.6
Kimi K2.6
Kimi K2.6 family · Provider: moonshot-ai · Status: active · Released: 2026-04-20
Kimi K2.6 is a Kimi K2.6 model developed by moonshot-ai, a Chinese AI company headquartered in 13F, Building 1, JD Technology Building, 76 Zhichun Road, Haidian District, Beijing, China, released 2026-04-20 with a 262,144-token context window and open weights.
| Key fact | Value |
|---|---|
| Model ID | kimi-k2.6 |
| Architecture | Mixture-of-Experts (MoE): 1T total / 32B activated; 384 routed experts (8 selected + 1 shared); MLA attention, SwiGLU activation; 61 layers (1 dense); MoonViT vision encoder (400M) |
| Context window | 262,144 tokens |
| Max output | 98,304 tokens |
| Open weights | Yes |
| License | Modified MIT |
| Self-hosting | Yes |
| API available | Yes |
| API pricing | $0.95 input / $4 output per 1M tokens (USD) · provider pricing page |
| Cloud providers | Moonshot AI Platform |
Capabilities
| Capability | Supported |
|---|---|
| Reasoning | Yes |
| Coding | Unknown |
| Math | Unknown |
| Chinese | Unknown |
| English | Unknown |
| Multilingual | Unknown |
| Vision | Yes |
| Audio | Unknown |
| Video | Unknown |
| Tool calling | Unknown |
| Function calling | Unknown |
| Structured output | Unknown |
| Agent capability | Unknown |
| RAG | Unknown |
| Computer use | Unknown |
Benchmark results
| Benchmark | Version | Score | Metric | Date | Source type | Source |
|---|---|---|---|---|---|---|
| HLE | — | 36.4 (text-only, no tools) | accuracy | 2026-04 | vendor_reported | link |
| HLE | — | 55.5 (with tools) | accuracy | 2026-04 | vendor_reported | link |
Benchmark scores are single data points, not universal rankings. Vendor-reported scores are labeled as such.
Known limitations
- Maximum output of 98,304 tokens is the documented max generation length in the official model card's evaluation configuration
Kimi K2.6 is Moonshot’s visual + text model with thinking and non-thinking modes for dialogue and agent tasks, in a 256K context window. Pricing is $0.95 input / $4.00 output per 1M tokens (cache hits $0.16, as of 2026-09-20).
Data interpretation
| Field | Value | Evidence type |
|---|---|---|
| Context window | 262,144 tokens | Official |
| Architecture | Mixture-of-Experts (MoE): 1T total / 32B activated; 384 routed experts (8 selected + 1 shared); MLA attention, SwiGLU activation; 61 layers (1 dense); MoonViT vision encoder (400M) | Official |
| Open weights | Yes | Official |
| License | Modified MIT | Official |
| API pricing | $0.95 / $4 per 1M tokens (USD) | Official |
| HLE | 36.4 (text-only, no tools) accuracy | Vendor-reported |
| HLE | 55.5 (with tools) accuracy | Vendor-reported |
| Family position | 1 model in the Kimi K2.6 family | China AI Hub analysis |
Evidence types: Official = vendor documentation, pricing pages or model cards. Vendor-reported = benchmark scores published by the vendor. China AI Hub analysis = derived from the database itself. See the sourcing policy.
Sources
Confidence and source hierarchy per the sourcing policy. Facts change; check the source before relying on this page.
What is the context window of Kimi K2.6?
Kimi K2.6 has a 262,144-token context window and a maximum output of 98,304 tokens.
Is Kimi K2.6 open weight?
Yes — Kimi K2.6 weights are openly available under the Modified MIT license.
How much does Kimi K2.6 cost through the API?
$0.95 per 1M input tokens and $4 per 1M output tokens (USD).
Where does China AI Hub get its Kimi K2.6 data?
From 3 sources (official pages first), last verified 2026-09-27. Benchmark scores are labeled by source type; see the sourcing policy for details.