Kimi K2.6
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.
kimi-k2.6Released 20 Apr 2026 (Apr 2026)Pricing
List prices at Moonshot AI, in US dollars per million tokens.
| Tier | Input / 1M | Output / 1M | |
|---|---|---|---|
| Standard | $0.95 | $4.00 | |
| Cached input 83% off input | $0.16 | 83% off input |
What it costs a month
At Moonshot AI list prices, 30 days.
- Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$514
- RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$410
- Coding agent1,000 requests a day, 40K in / 2K out, cached$716
- Overnight batch20,000 requests a day, 4K in / 0.4K out$3,240
Where to run it
20 hosts serve Kimi K2.6. The comparison is on blended price, 3 parts input to 1 part output.
| Host | In / out per 1M | Model id | |
|---|---|---|---|
Moonshot AILab API kimi-k2.6 | $0.95 / $4.00 | kimi-k2.6 | |
CLF AI GatewayCloud kimi-k2.6 | $0.38 / $1.60 −60% vs headline | kimi-k2.6 | |
Kyma APICloud kimi-k2.6 | $0.591 / $2.76 −34% vs headline | kimi-k2.6 | |
W&B InferenceInference Kimi-K2.6 | $0.65 / $3.41 −22% vs headline | Kimi-K2.6 | |
DeepInfraInference Kimi-K2.6 | $0.75 / $3.50 −16% vs headline | Kimi-K2.6 | |
Novita AIInference kimi-k2.6 | $0.80 / $3.40 −15% vs headline | kimi-k2.6 | |
GMI CloudInference Kimi-K2.6 | $0.855 / $3.60 −10% vs headline | Kimi-K2.6 | |
EmpirioLabs AICloud kimi-k2-6 | $0.894 / $3.71 −7% vs headline | kimi-k2-6 | |
AIHubMixGateway kimi-k2.6 | $0.95 / $4.00 | kimi-k2.6 | |
Azure AI FoundryCloud kimi-k2.6 | $0.95 / $4.00 Retires 16 Apr 2027 | kimi-k2.6 | Retires 16 Apr 2027 |
kimi-k2.6 | $0.95 / $4.00 | kimi-k2.6 | |
StreamLakeCloud kimi-k2.6 | $0.95 / $4.00 | kimi-k2.6 | |
BasetenInference Kimi-K2.6 | $0.95 / $4.00 | Kimi-K2.6 | |
Nebius AI StudioInference Kimi-K2.6 | $0.95 / $4.00 | Kimi-K2.6 | |
ReplicateInference kimi-k2.6 | $0.95 / $4.00 | kimi-k2.6 | |
Fireworks AIInference kimi-k2p6 | $0.95 / $4.00 | kimi-k2p6 | |
TensormeshCloud Kimi-K2.6 | $0.96 / $4.00 | Kimi-K2.6 | |
SailCloud Kimi-K2.6 | $1.00 / $4.00 +2% vs headline | Kimi-K2.6 | |
Together AIInference Kimi-K2.6 | $1.20 / $4.50 +18% vs headline | Kimi-K2.6 | |
NVIDIA NIMInference kimi-k2.6 | - | kimi-k2.6 |
Show 1 more listings: regions, snapshots and variantsHide regional listings
| Fireworks AI | |||||||
| Alias kimi-k2p6 | $0.95 / $4.00 | kimi-k2p6 | |||||
Capabilities
Core
- Streaming
- Function calling
- JSON mode
- Structured outputs
Multimodal
- Vision
- Image input
- Audio
- Image generation
- Video
Advanced
- Reasoning
- Tool calling
- MCP
- Prompt caching
- Batch API
- Embeddings
- Takes in
- TextImagesVideo
- Produces
- Text
- Features
- Video input
- Open weights
- License: Modified MIT LicenseRuns in about 2464.5 GB of GPU memoryRuntimes: SGLang, vLLM
Cheaper alternatives
Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.
Same model, cheaper host
- CLF AI Gateway60% lesskimi-k2.6$0.38 / $1.60
- Kyma API34% lesskimi-k2.6$0.591 / $2.76
- W&B Inference22% lessKimi-K2.6$0.65 / $3.41
- DeepInfra16% lessKimi-K2.6$0.75 / $3.50
- Novita AI15% lesskimi-k2.6$0.80 / $3.40
Smaller Moonshot AI models
No cheaper Moonshot AI model keeps its key capabilities.
Other frontier labs
- 1% less
- 12% less
- 27% less
- 88% less
Open-weight picks
- 60% less
- 63% less
- 69% less
- 72% less
Model line
Newer and older versions, at each lab's own price.
Call it
Using Moonshot AI's API.
from litellm import completion
response = completion(
model="moonshot/kimi-k2.6",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)Quick answers
How much does Kimi K2.6 cost?
On Moonshot AI, Kimi K2.6 costs $0.95 per million input tokens and $4.00 per million output tokens. Cached input costs $0.16 per million tokens. The cheapest host is CLF AI Gateway, at $0.38 per million input tokens and $1.60 per million output tokens.
Which APIs offer Kimi K2.6?
Kimi K2.6 runs on 20 APIs: Moonshot AI, CLF AI Gateway, Kyma API, W&B Inference, DeepInfra, Novita AI, GMI Cloud, EmpirioLabs AI, AIHubMix and 11 more.
What is Kimi K2.6's context window?
262K tokens, with up to 262K tokens in one reply.
Is Kimi K2.6 being retired?
Azure AI Foundry retires it on 16 Apr 2027. 19 hosts keep serving it.
Is there a newer version of Kimi K2.6?
Yes. Kimi K3.1 is the newest model in this line.
What are cheaper alternatives to Kimi K2.6?
Models that cost less and keep most of what it can do: GPT-5.4 Mini on OpenAI (1% less), Gemini 3.8 Flash on Google AI Studio (12% less) and Kimi K2.7 Code on CLF AI Gateway (60% less). Kimi K2.6 itself costs 60% less on CLF AI Gateway. Savings are on a mix of 3 input tokens to 1 output token.
When was Kimi K2.6 released?
Moonshot AI released it on 20 Apr 2026.
What is Kimi K2.6's API model id?
Use kimi-k2.6 with Moonshot AI. Other hosts use their own ids: Kimi-K2.6 on W&B Inference, Kimi-K2.6 on DeepInfra, Kimi-K2.6 on GMI Cloud, kimi-k2-6 on EmpirioLabs AI.
Sources
- Baseten pricing
- DeepInfra pricing
- GMI Cloud pricing
- Nebius AI Studio pricing
- Sail pricing
- Tensormesh pricing
- Together AI pricing
- W&B Inference pricing
- EmpirioLabs AI pricing
- AIHubMix pricing
- Azure AI Foundry pricing
- CLF AI Gateway pricing
- Moonshot AI pricing
- Novita AI pricing
- NVIDIA NIM pricing
- Replicate pricing
- StreamLake pricing
- Fireworks AI pricing
- LiteLLM model map
Raw catalog entry for kimi-k2.6 on Moonshot AI
{
"cache_read_input_token_cost": 1.6e-7,
"input_cost_per_token": 9.5e-7,
"litellm_provider": "moonshot",
"max_input_tokens": 262144,
"max_output_tokens": 262144,
"mode": "chat",
"output_cost_per_token": 0.000004,
"source": "https://platform.kimi.ai/docs/pricing/chat-k26",
"supports_function_calling": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_video_input": true,
"supports_vision": true
}