Experiential
ModelsLogsInsights
Star us on GitHubDocsSettings
Sign in
Continue with GoogleContinue with GitHub
or
Models

Kimi K3

kimi-k3by KimirecommendedZDR enforceable

Kimi K3 is a 2.8T parameter 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...

CompareOpen in Playground

Context

1.05M

Max output

—

Input from

$3 / M

Output from

$12.75 / M

Fastest

3474 tok/s

Released

Jul 2026

Waterfall

Status
Fireworks via Experiential Cloudexperiential———1.05M—$3$15cache$0.30—ZDRactiveestimated
Azure Foundry via Experiential CloudexperientialFW-Kimi-K3——100%1.05M—$3.30$16.50cache$0.33—retainsactivemeasured
TTencent Cloud via Experiential Cloudexperiential———1.05M—$3$15cache$0.30—retainsactive—
WWafer via Experiential Cloudexperiential———1.05M—$3$12.75cache$0.30—retainsactive—
Azure FoundryKimi-K3——100%1.05M—————retainsactivemeasured
Azure FoundryKimi-K3-2026-07-29deployed 2026-07-29——100%1.05M—————retainsactivemeasured

Supported parameters

primary route

Sampling

Stop sequences

Tools & structure

Parallel tool calls*

Streaming

Streaming

Limits

Max-tokens fieldmax_tokens

* not supported on every route — the gateway adapts or drops it (with a warning) on stricter providers.

Shown for the primary route; fallback routes may differ. Sending an unsupported field? See error reference.

Azure OpenAI docs

Quickstart

I want you to route my LLM calls for "kimi-k3" through the Experiential gateway instead of
calling the provider directly. It speaks the OpenAI Chat Completions API, so this is a base-URL
and key swap. Please:

1. Point the client at https://api.experientiallabs.ai/v1 as the base URL.
2. Authenticate with my Experiential API key from the EXPLABS_API_KEY environment variable. If
   it isn't set, stop and tell me to create one under Settings -> API keys and export it.
3. Use the model id "kimi-k3" exactly.
4. Update every place my code builds an LLM client for this model to use that base URL and key,
   leaving streaming and tool-calls as they are.
5. Make one test call and show me the reply plus the token usage, so we confirm it runs on my
   Experiential credits.

Tell me which files you changed.

Set EXPLABS_API_KEY to an organization API key before running your agent.

Benchmarks

Hugging Face
  • MMLU-Propublic leaderboard · Aug 202689.3%
  • GPQA DiamondHugging Face · Aug 202693.5%
  • SWE-bench Verifiedpublic leaderboard · Aug 202693.4%
  • LiveBenchpublic leaderboard · Aug 202679.2%
  • AIME 2025public leaderboard · Aug 202695.8%
  • AIME 2026vendor reported · Aug 202696.7%
  • Humanity's Last ExamHugging Face · Aug 202643.5%
  • LMArena EloLMArena · Aug 20261476
  • Terminal-BenchHugging Face · Aug 202688.3