Kimi K3 API

Kimi K3 is a large language model by Moonshot AI, available on the Venice API as kimi-k3. Private and end-to-end encrypted variants are available.Kimi K3 is an ultra-large-scale, open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at navigating large repositories, using tools, debugging, and iterating against images, logs, tests, and runtime feedback.

Kimi K3 API pricing

Model ID	Variant	Privacy	Price
`kimi-k3`	Standard	Private	$3.75 input /$ 18.75 output per 1M tokens
`kimi-k3-fast-api`	Fast	Private	$4.50 input /$ 22.50 output per 1M tokens
`e2ee-kimi-k3-p`	E2EE	End-to-end encrypted	$3.75 input /$ 18.75 output per 1M tokens

Kimi K3 specifications

Spec	Value
Provider	Moonshot AI
Released	Jul 16, 2026
Privacy	Private, End-to-end encrypted
Open weights	Yes
Context window	1M tokens
Max output	128K tokens
Reasoning effort	none, low, high, max

How to use the Kimi K3 API

Send requests to POST https://api.venice.ai/api/v1/chat/completions with "model": "kimi-k3" and your API key.

curl https://api.venice.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $VENICE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kimi-k3",
    "messages": [{ "role": "user", "content": "Explain TEE attestation in two sentences." }],
    "reasoning_effort": "max"
  }'

Kimi K3 API FAQ

How much does the Kimi K3 API cost?

3.75 per 1M input tokens and

18.75 per 1M output tokens, with cached input at

0.38 per 1M. The Fast variant costs

4.50 input and $22.50 output. Prices are in USD and can be paid in DIEM at parity.

What is the Kimi K3 model ID?

Use kimi-k3 as the model parameter. Other variants: kimi-k3-fast-api (Fast), e2ee-kimi-k3-p (E2EE).

Is the Kimi K3 API private?

The Standard and Fast variants are private: requests run on infrastructure Venice controls with zero data retention, and prompts and outputs are never stored or used for training. The E2EE variant is end-to-end encrypted: prompts are encrypted on your device and decrypted only inside an attested hardware enclave, so neither Venice nor the GPU provider can read them.

What is the context window of Kimi K3?

1M tokens of context, with up to 128K output tokens per response.

What does Kimi K3 support?

Kimi K3 supports function calling, structured outputs, reasoning, image input, web search and prompt caching. Reasoning effort is adjustable with reasoning_effort: none, low, high and max (default max).

Which endpoint does the Kimi K3 API use?

Call POST /chat/completions. /responses (Alpha) is also supported.

Related models

Kimi K2.6 API: $0.75 input /$ 3.50 output per 1M tokens
Kimi K2.5 API: $0.56 input /$ 3.50 output per 1M tokens
Claude Sonnet 5.5 API: $3.75 input /$ 18.75 output per 1M tokens
GPT-5.4 API: $3.13 input /$ 18.80 output per 1M tokens
Claude Sonnet 4.6 API: $3.60 input /$ 18.00 output per 1M tokens
Claude Sonnet 5 API: $3.00 input /$ 15.00 output per 1M tokens

Explore

By modality