Kimi K3 - Kimi API Platform

5 min read Original article ↗

Introducing Kimi K3

Kimi K3 is Kimi’s most capable flagship model to date, with 2.8 trillion parameters. It is built on Kimi Delta Attention (KDA), a hybrid linear attention mechanism, and Attention Residuals, with native visual understanding and a 1M-token context window. It is the world’s first open-source model in the 3-trillion-parameter class, designed for frontier intelligence scenarios including long-horizon coding, knowledge work, and reasoning. For complete benchmarks and case studies, see the technical blog. Kimi is currently working closely with inference partners and open-source maintainers to align technical details and ensure the model launches reliably across the ecosystem. The full model weights will be released by July 27, 2026. More details on architecture, training, and evaluation will be published with the Kimi K3 technical report.

A 3-trillion-scale open-source model

Kimi K3 is the first open-source model to reach 2.8 trillion parameters. This is the latest step in Kimi’s continued push of model-scale boundaries: in 9 of the past 12 months (2025/07–2026/07), Kimi models have maintained the frontier in open-source model scale. Open-source frontier model scale over time Kimi K3 is built on Kimi Delta Attention (KDA) and Attention Residuals (AttnRes). Both architectural updates are designed to help information flow more smoothly through longer sequences and deeper models. We also further increased the sparsity of the Mixture of Experts (MoE): with the Stable LatentMoE framework, the model efficiently activates 16 out of 896 experts. Together with improvements in training methodology and data recipes, these structural advances give Kimi K3 roughly 2.5x the overall scaling efficiency of K2, converting compute into capability more effectively. Kimi K3 architecture

Coding

Kimi K3 has strong long-horizon coding capabilities. With minimal human supervision, it can sustain long-running engineering tasks, understand and work with large codebases, and coordinate terminal tools. Kimi K3 also excels at tasks that combine software engineering and visual reasoning. It can use screenshots and visual feedback to improve workflows in game development, frontend engineering, CAD, and related scenarios.

Knowledge work

Kimi K3 advances end-to-end knowledge work. Beyond public benchmarks, Kimi K3 (max) also shows consistent gains in our internal evaluations. These evaluations reflect recurring task patterns and challenges from real user-agent collaboration workflows. Kimi K3 demonstrates consistent advantages across production-oriented workflows, indicating broad improvements in agentic knowledge-work capabilities.

Access requirements

Kimi K3 is a flagship model: it is unlocked after a successful top-up (minimum $1). Your cumulative top-up amount also determines your account tier and rate limits (concurrency, RPM, TPM, TPD) — see Recharge and Rate Limits.

Get started

The examples require Python 3.9+ and the OpenAI SDK. Install the SDK and initialize the client once; later Python examples reuse client.

Basic call

Reasoning effort

K3 always has thinking mode enabled and supports configuring its reasoning effort with the top-level reasoning_effort request field.

Streaming

Streaming responses provide separate reasoning_content and final-answer content deltas. See Streaming Output for details.

Vision input

For vision messages, content must be an array of objects, not a serialized string. See Vision Input for formats and limits.

Structured output

Use json_schema with strict: true to constrain the final message.content. Parse only that field, not reasoning_content.

See Structured Output.

Partial Mode

Add an assistant message with partial=True at the end of messages to continue from a text prefix. Prepend that prefix when displaying the final result.

See Partial Mode.

Use tool_choice="required" on the first turn to require at least one tool call. After executing every call, return the complete assistant message and append one tool result with the matching tool_call_id for each call.

See Tool Choice.

Place a complete tool definition in a system message without content. The tool becomes available from that message onward.

  • Include the complete name, description, and parameters definition.
  • The declaration takes effect at its position in messages.
  • Keep this message in later request history; the server does not retain it.

See Dynamic Tool Loading.

1M context and automatic caching

Context caching is automatic for regular model requests; no cache ID, TTL, or extra parameter is required. Keep the long prefix unchanged so later requests can automatically attempt a cache hit.

See Context Caching.

Official tools are integrated through Formula:

  1. Fetch tool definitions from the Formula /tools endpoint.
  2. Add those definitions to the Chat Completions tools field.
  3. When the model returns tool_calls, submit each function name and arguments to the Formula /fibers endpoint.
  4. Add the complete assistant message and Fiber output as the corresponding tool message.
  5. Call Chat Completions again until the model returns a final answer.

See Official Tools for the complete client and API contract. Web search is being updated and is not recommended for use in the near term.

Important limits

  • Reasoning effort is configured with the top-level reasoning_effort request field and supports low, high, and max (default max); K3 always has thinking mode enabled.
  • max_completion_tokens defaults to 131072 and can be set up to 1048576.
  • temperature=1.0, top_p=0.95, n=1, presence_penalty=0, and frequency_penalty=0 are fixed; omit them from requests.
  • Return the complete assistant message unchanged in multi-turn conversations and tool calls.
  • Vision input does not support public image URLs. Use base64 or ms://<file-id>, and make content an array of objects.
  • Web search is being updated and is not recommended for production workflows in the near term.

FAQ

Model Pricing

For token pricing details, refer to Model Pricing.