AI Directory : AI Agent, AI Chatbot, AI Code Assistant, Large Language Models (LLMs)
What is Kimi K3.1 Pro?
Kimi K3.1 Pro is the first model in Moonshot AI's K3.1 series, released on October 24, 2026. It is a post-trained version of Kimi K3, described by Moonshot as its most capable 3T-class open-weight model, with a 1-million-token context window. It targets developers, analysts and teams that run long coding, 3D modeling or financial research tasks through chat or the API.
How to Use Kimi K3.1 Pro
The model is available on Kimi.com, Kimi Work, Kimi Code and the Kimi API. In the Kimi web app, start a new chat or pick an agent mode such as Slides, Deep Research, Docs, Sheets, Build or 3D Make, and ask for the task.
For the API, create a key on the Kimi API Platform and call the model ID kimi-k3.1-pro, or kimi-k3.1-pro-highspeed for faster responses. Moonshot says the full model weights will be released by November 4, 2026, for teams that want to self-host.
Key Features
- 1M-token context All K3 series models on the API support a 1,048,576-token window.
- Long-horizon coding 65.7% on Terminal-Bench 4.0 and 61.2% on Moonshot's internal Kimi Code Bench 3.0 Long-horizon, up from 55.95% for K3.
- 3D creation 92.7% on BenchCAD with tools; it can script Blender or CAD tools, and 3D Make is live on kimi.com.
- Finance workflows 64.4 on OfficeQA Pro v2, plus fewer factual errors on Moonshot's internal hallucination stress test.
- Highspeed variant K3.1 Pro Highspeed runs 2-3x faster with the same context window.
Why Choose Kimi K3.1 Pro?
It pairs open weights with a 1M-token context, and its API price sits below the previous K3: $2 versus $3 per million input tokens and $10 versus $15 per million output tokens. Moonshot also reports that it removed cyber-related training data and applied training controls to reduce misuse risk while keeping general usefulness. Benchmark figures above are Moonshot's own.
Use Cases and Applications
Developers can run multi-step coding agents across large repositories or build games in Unity, Unreal Engine and Godot. Product and furniture designers can generate parametric 3D models from a sketch or text brief. Analysts can build linked financial models and source-grounded research reports in Kimi's Sheets and Docs modes.
Frequently Asked Questions
How much does Kimi K3.1 Pro cost?
On the API it costs $2.00 per million input tokens, $10.00 per million output tokens and $0.20 per million cached input tokens; Highspeed is $4.00 and $20.00. Batch inference is billed at 60% of the standard price. On Kimi.com there is a free plan, and paid membership starts with Plus at $15/month billed annually ($19 month to month); Highspeed access is listed from the Pro plan.
Is Kimi K3.1 Pro open source?
Moonshot calls it open-weight and says the full weights will be released by November 4, 2026.
What is the difference between K3.1 Pro and K3.1 Pro Highspeed?
They share the same capabilities and 1M-token context; Highspeed is 2-3x faster and costs twice as much per input and output token.
Can requests fail at launch?
Kimi's pricing page warns that K3.1 Pro requests may queue or fail at launch, with wait times depending on overall load.