Requesty
Nebius AI logo
Nebius AI

moonshotai/kimi-k2.6

Kimi K2.6 is an open-source, native multimodal agentic model built through continual pretraining on approximately 15 trillion mixed visual and text tokens atop Kimi-K2-Base.

VisionReasoningTool callingJSON schema
Try moonshotai/kimi-k2.6CompareDocs

Specifications

Context window256K tokens
Max output128K tokens
API typechat
AddedJun 16, 2026
Model IDnebius/moonshotai/kimi-k2.6
Data retentionNo
Used for trainingNo
Provider location🇪🇺 EU
Privacy policyNebius Privacy Policy→

Benchmarks

Released 2026-04-20
Coding Indexcoding
47.1%

Artificial Analysis Coding Index — a composite of coding evaluations including LiveCodeBench, SciCode and Terminal-Bench.

GPQA Diamondreasoning
91.1%

Graduate-level physics, chemistry & biology questions designed to resist Googling.

Intelligence Indexreasoning
42.8%

Artificial Analysis Intelligence Index — a composite of multiple evaluations measuring overall model capability.

Scores are sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and do not capture every aspect of model quality. Always test on your own workload.

Pricing

Input / 1M
$0.95
Output / 1M
$4.00
Cache write
N/A
Cache read
N/A
Estimated cost
100K input + 10K output$0.14
1M input + 100K output$1.35
10M input + 1M output$13.50

Requesty charges exactly what the upstream provider charges, no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to nebius/moonshotai/kimi-k2.6.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="nebius/moonshotai/kimi-k2.6",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other Nebius AI models

glm-5.2
1M ctx$1.40/1M
Coding 51
gemma-3-27b-it
128K ctx$0.10/1M
qwen/qwen3-32b
128K ctx$0.10/1M
qwen/qwen3-next-80b-a3b-thinking
128K ctx$0.15/1M
nousresearch/hermes-4-70b
128K ctx$0.13/1M
qwen/qwen3-30b-a3b-instruct-2507
128K ctx$0.10/1M

Frequently asked questions

How much does moonshotai/kimi-k2.6 cost?
moonshotai/kimi-k2.6 is priced at $0.95 per million input tokens and $4.00 per million output tokens when accessed via Requesty. Requesty charges exactly what the upstream provider charges, we don't add markup.
What is the context window of moonshotai/kimi-k2.6?
moonshotai/kimi-k2.6 has a context window of 256K tokens, with a maximum output of 128K tokens per response. That's roughly 341 words of input you can fit in a single prompt.
How does moonshotai/kimi-k2.6 perform on benchmarks?
moonshotai/kimi-k2.6 scores 95.9% on τ²-Bench, 91.1% on GPQA Diamond, 53.5% on SciCode. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can moonshotai/kimi-k2.6 do?
moonshotai/kimi-k2.6 supports vision input, tool calling, extended reasoning, structured outputs (JSON schema). You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use moonshotai/kimi-k2.6 with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "nebius/moonshotai/kimi-k2.6". The Quickstart above shows Python, JavaScript and cURL snippets.
Can I run moonshotai/kimi-k2.6 through Requesty?
Yes. moonshotai/kimi-k2.6 runs through Requesty's OpenAI-compatible API, served from Nebius AI. You do not host the model yourself: point base_url at Requesty, set the model to "nebius/moonshotai/kimi-k2.6", and requests are routed to the upstream provider with automatic failover. The same key gives you 400+ other models too.

Access moonshotai/kimi-k2.6 through Requesty

One API key, 400+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

Get started freeAll Nebius AI models