Moonshot AI
Operational· checked every few minutes

Kimi K2.6 API

The Kimi K2.6 API for coding agents and long-running tasks, 68% below Moonshot's list price

Price per 1M tokens68%off
Input
$0.30$0.95
Output
$1.30$4.00
Cached input
$0.05
Get your API keyTry it in the playground
Model IDkimi-k2.6
chat.py
from openai import OpenAI
client = OpenAI(
    base_url="https://api.zurelay.com/v1",
    api_key="YOUR_ZURELAY_KEY",
)
stream = client.chat.completions.create(
    model="kimi-k2.6",
    messages=[{"role": "user", "content": "Explain quicksort in two sentences."}],
    stream=True,
)
for chunk in stream:
    if chunk.choices:
        print(chunk.choices[0].delta.content or "", end="")

Base URL https://api.zurelay.com/v1. Works with any OpenAI SDK. Chat API docs

Context window
256K tokens
Reasoning
none · high
Input
Text, Image, Video
Released
Apr 20, 2026
Status
Operational

Pricing

Kimi K2.6 API pricing. A fraction of the list price.

Pay as you go from prepaid credit, with no subscription. Top up from $10. Requests that fail are never billed.

RateZurelayMoonshot AI list
Input
per 1M tokens
$0.30$0.95
Output
per 1M tokens
$1.30$4.00
Cached input
per 1M tokens
$0.05—

One rate at every prompt length, up to the full 256K context window. Streaming, tool calls and structured outputs cost nothing extra.

Every Kimi model and price
Savings calculator

Move the sliders to your monthly usage.

Input tokens per month50M
Output tokens per month10M
Zurelay
$28.00
Moonshot AI list price
$87.50
You save every month$59.50

$714.00 a year

Overview

What is Kimi K2.6?

Kimi K2.6 is Moonshot AI's open-weight K2 model for long-horizon coding, front-end design and agents, released in April 2026, with a 262K-token context window and image and video input. The Zurelay Kimi K2.6 API serves it through one OpenAI-compatible endpoint at 68% below Moonshot's list price.

Moonshot AI released Kimi K2.6 in April 2026 and published the weights on Hugging Face under a modified MIT license. It is a Mixture-of-Experts model with 1T total parameters and 32B active per token, routing each token to 8 of 384 experts, with a 400M-parameter vision encoder, MoonViT. It holds 262,144 tokens of context. On Zurelay the model ID is kimi-k2.6, and it costs $0.30 per 1M input tokens and $1.30 per 1M output tokens.

Moonshot built K2.6 for long-horizon coding across languages such as Rust, Go and Python, and across front-end, DevOps and performance work. In one of its tests the model spent 13 hours optimizing exchange-core, an open-source financial matching engine, trying 12 strategies over more than 1,000 tool calls. It also turns short prompts into complete front-end interfaces, and Moonshot reports that its agent swarms reach 300 sub-agents over 4,000 coordinated steps, up from 100 and 1,500 on K2.5.

On Moonshot's own tests in thinking mode, K2.6 scores 58.6% on SWE-Bench Pro (GPT-5.4: 57.7%, Claude Opus 4.6: 53.4%), 80.2% on SWE-Bench Verified, 66.7% on Terminal-Bench 2.0 and 54.0 on Humanity's Last Exam with tools. Kimi K2.5 scored 50.7%, 76.8%, 50.8% and 50.2 on the same tests. These are Moonshot's numbers, not independent ones.

Thinking is on by default, as on Moonshot's own API, and the reasoning comes back as reasoning_content. On Zurelay, set reasoning_effort to none for quick, direct answers without thinking; any other level turns it on. K2.6 reads text, images and video and writes text. It calls tools, follows a JSON schema in response_format and streams. In our tests it wrote about 55 tokens a second and found one fact buried in a 198K-token prompt. You pay $0.30 per 1M input tokens and $1.30 per 1M output tokens against Moonshot's list price of $0.95 and $4.00, with cached input at $0.05.

Strengths

Where Kimi K2.6 shines. And what teams build with it.

01

Long-horizon coding

58.6% on SWE-Bench Pro and 66.7% on Terminal-Bench 2.0 in Moonshot's tests, and one 13-hour optimization run of more than 1,000 tool calls.

02

Images and video in

Native vision from a 400M-parameter encoder: send screenshots, designs, charts or video clips next to your text.

03

Thinking when you want it

Thinks by default for hard problems. Set reasoning_effort to none for fast, direct answers that use fewer output tokens.

04

Open weights

Moonshot published the weights under a modified MIT license, so prompts you build on Zurelay carry over to a self-hosted copy later.

Use cases

  • Coding agents

    Multi-file changes, test loops and terminal work in agent harnesses, over long sessions.

  • Front-end from a brief or a design

    Turn a short prompt or a screenshot of a design into a working interface, then send screenshots back to refine it.

  • Research agents

    Search and browse over many steps. Moonshot reports 83.2% on BrowseComp.

  • Video and documents

    Ask questions about video clips, scanned pages and charts in the same request as your text.

Get started

Call Kimi K2.6 in three steps.

No waitlist and no new SDK. If your code already talks to OpenAI, it already talks to Zurelay.

  1. 1

    Create a key

    Sign up, add credit and create an API key. Set a monthly budget or a rate limit per key if you like.

    Get your API key
  2. 2

    Point your SDK at Zurelay

    Change the base URL. Everything else in your code stays the same.

    https://api.zurelay.com/v1
  3. 3

    Use kimi-k2.6

    Send chat completions as usual. Streaming, tool calls and usage reporting work as you expect.

    kimi-k2.6

Works with the tools you already use

Compare

Similar models. All on the same key.

60%off

Kimi K3.1 Pro

$0.80 / $4.00 per 1M tokens

Pick Kimi K3.1 Pro, Moonshot's newest flagship, for the hardest coding, 3D and financial work and a 1M-token context, at a higher price.

Kimi K3.1 Pro API
68%off

Kimi K2.5

$0.20 / $0.95 per 1M tokens

Pick Kimi K2.5, the previous K2 model, for the same 262K context and image and video input at a lower price.

Kimi K2.5 API
73%off

GLM-5.3

$0.38 / $1.21 per 1M tokens

Pick GLM-5.3, Z.ai's flagship, for text-only coding agents that need a 1M-token context and up to 128K output tokens.

GLM-5.3 API

FAQ

Kimi K2.6 API questions.

Anything else? Ask our team and we’ll answer by email.

How much does the Kimi K2.6 API cost on Zurelay?

Kimi K2.6 costs $0.30 per 1M input tokens and $1.30 per 1M output tokens on Zurelay, with cached input at $0.05. Moonshot's list price is $0.95 input and $4.00 output, so you save 68%. Reasoning tokens count as output, and one rate applies at every prompt length.

How do I call the Kimi K2.6 API with the OpenAI SDK?

Install the official OpenAI SDK, set the base URL to https://api.zurelay.com/v1 and use your Zurelay API key. Then pass model "kimi-k2.6" to chat.completions.create; Zurelay also accepts moonshotai/kimi-k2.6. Moonshot recommends temperature 1.0 with thinking on and 0.6 with it off, and top_p 0.95.

What is the Kimi K2.6 context window?

Kimi K2.6 holds 262,144 tokens of context, and in our tests it found one fact buried in a 198K-token prompt. It can write long answers. Reasoning and the answer share the output budget you set with max_tokens, so leave headroom on hard prompts.

Can Kimi K2.6 read images and video?

Yes. Send images in image_url parts and video in video_url parts, as https links or base64 data URLs. Images and video count as input tokens. Zurelay doesn't take PDFs or audio for Kimi K2.6: a request that includes one is turned away with a clear error before anything is billed.

Can I turn off Kimi K2.6's thinking?

Yes. Thinking is on by default, as on Moonshot's own API, and the reasoning comes back in reasoning_content. Set reasoning_effort to none for direct answers without thinking, which come back faster and use fewer output tokens. Any other level turns thinking on.

Does Kimi K2.6 support tool calling and JSON output?

Yes. Pass tools as you would to any OpenAI model. Moonshot's own API takes only tool_choice auto or none while thinking is on; on Zurelay a tool forced by name is still called, and tool_choice "none" works. For JSON, send response_format with json_object or json_schema and the answer comes back as JSON.

How is Kimi K2.6 different from Kimi K2.5?

K2.6 is the newer K2 model, with the same size, the same 262K context and the same image and video input. Moonshot reports clear gains in coding and agent work: 58.6% against 50.7% on SWE-Bench Pro, 66.7% against 50.8% on Terminal-Bench 2.0 and 50.0 against 27.8 on Toolathlon. K2.5 costs less per token.

Should I use Kimi K2.6 or Kimi K3.1 Pro?

Kimi K3.1 Pro is Moonshot's newer flagship, built on Kimi K3, with a 1M-token context, always-on thinking and stronger coding, 3D and finance results, at a higher price. Pick K2.6 when cost matters, when you need video input, or when you want answers without thinking.

Is Kimi K2.6 on Zurelay the same model Moonshot serves?

Yes. Requests go to kimi-k2.6 itself, Moonshot's open-weight model. Zurelay doesn't swap in a smaller or substitute model. Only smart routing, which you turn on yourself, can answer with a close alternative when the model is down, and the x-zurelay-model header names the model that answered. Output is sampled, so wording varies from run to run, just as it does on Moonshot's own API. Zurelay is an independent service and is not affiliated with Moonshot AI.

More models on Zurelay

All models

Stop paying list price.
Start saving today.

Create a key in seconds, point the SDK you already use at Zurelay, and every request costs up to 90% less from the first token.

Get your API keyTalk to an engineer