impulse

Kimi K2 Instruct

Open weights
API id Kimi-K2-Instruct
Input / 1M
$0.50
at DeepInfra
Output / 1M
$2.00
at DeepInfra
Context window
131K
tokens
Max output
131K
tokens per reply
Hosts
8
10 listings
Released
-
date not published

Pricing

List prices at DeepInfra, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.50$2.00

What it costs a month

At DeepInfra list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out$300
  • RAG over documents2,000 requests a day, 8K in / 0.6K out$312
  • Coding agent1,000 requests a day, 40K in / 2K out$720
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$1,680
Use your own numbers

Where to run it

8 hosts serve Kimi K2 Instruct. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
DeepInfraInference
Kimi-K2-Instruct
$0.50 / $2.00
0905
Kimi-K2-Instruct-0905
$0.571 / $2.29
+14% vs headline
GMI CloudInference
0905
Kimi-K2-Instruct-0905
$0.57 / $2.29
+14% vs headline
Novita AIInference
kimi-k2-instruct
$0.57 / $2.30
+15% vs headline
BasetenInference
0905
Kimi-K2-Instruct-0905
$0.60 / $2.50
+23% vs headline
Fireworks AIInference
kimi-k2-instruct
$0.60 / $2.50
+23% vs headline
Together AIInference
Kimi-K2-Instruct
$1.00 / $3.00
+71% vs headline
HyperbolicInference
Kimi-K2-Instruct
$2.00 / $2.00
+129% vs headline
Show 2 more listings: regions, snapshots and variants
DeepInfra
0905
Kimi-K2-Instruct-0905
$0.50 / $2.00
Fireworks AI
0905
kimi-k2-instruct-0905
$0.60 / $2.50
+23% vs headline

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
Text
Produces
Text
Open weights
License: Modified MIT License
Runtimes: 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

No host sells it for less than DeepInfra.

Smaller Moonshot AI models

No cheaper Moonshot AI model keeps its key capabilities.

Other frontier labs

  • Gemini 3.5 Flash Lite
    Google · on Google AI Studio
    $0.30 / $2.50
    Has every key capability
    3% less
  • GPT-5.4 Nano
    OpenAI · on OpenAI
    $0.20 / $1.25
    Has every key capability
    47% less
  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    77% less
  • GPT-6 Luna
    OpenAI · on OpenAI
    $0.10 / $0.50
    Has every key capability
    77% less

Open-weight picks

  • GLM 5.3
    Zhipu AI · on CLF AI Gateway
    $0.56 / $1.76
    Has every key capability
    2% less
  • Nemotron 3 Ultra 550B A55B
    NVIDIA · on Perplexity
    $0.25 / $2.50
    Has every key capability
    7% less
  • DeepSeek V4 Pro
    DeepSeek · on CLF AI Gateway
    $0.528 / $1.58
    Has every key capability
    9% less
  • Kimi K2.7 Code
    Moonshot AI · on CLF AI Gateway
    $0.38 / $1.60
    Has every key capability
    22% less

Call it

Using DeepInfra's API.

Python
from litellm import completion

response = completion(
    model="deepinfra/Kimi-K2-Instruct",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Kimi K2 Instruct cost?

On DeepInfra, Kimi K2 Instruct costs $0.50 per million input tokens and $2.00 per million output tokens.

Which APIs offer Kimi K2 Instruct?

Kimi K2 Instruct runs on 8 APIs: DeepInfra, StreamLake, GMI Cloud, Novita AI, Baseten, Fireworks AI, Together AI and Hyperbolic.

What is Kimi K2 Instruct's context window?

131K tokens, with up to 131K tokens in one reply.

Is Kimi K2 Instruct being retired?

No retirement date is scheduled for it on any API we track.

What are cheaper alternatives to Kimi K2 Instruct?

Models that cost less and keep most of what it can do: Gemini 3.5 Flash Lite on Google AI Studio (3% less), GPT-5.4 Nano on OpenAI (47% less) and GLM 5.3 on CLF AI Gateway (2% less). Savings are on a mix of 3 input tokens to 1 output token.

What is Kimi K2 Instruct's API model id?

Use Kimi-K2-Instruct with DeepInfra. Other hosts use their own ids: Kimi-K2-Instruct-0905 on StreamLake, Kimi-K2-Instruct-0905 on GMI Cloud, kimi-k2-instruct on Novita AI, Kimi-K2-Instruct-0905 on Baseten.

Sources

Raw catalog entry for Kimi-K2-Instruct on DeepInfra
{
  "max_input_tokens": 131072,
  "max_output_tokens": 131072,
  "input_cost_per_token": 5e-7,
  "output_cost_per_token": 0.000002,
  "litellm_provider": "deepinfra",
  "mode": "chat",
  "supports_tool_choice": true,
  "supports_function_calling": true
}