Kimi K3.1 Pro
NVIDIA's slimmed-down copy of Moonshot AI's Kimi K3.1 Pro, free to download and built to run on NVIDIA's newest graphics chips for developers building chatbots and AI agents.
Key facts
- Released
- Nov 11, 2026
- What it does
- Text and chat: Writes, answers questions and holds a conversation.
Can I use it?
In an app
No app mentioned so far.
For developers
No developer access mentioned so far.
Download
Free to download and run on your own computer: it has Open weights: The model's files are public, so anyone can download it, run it on their own computer and build on it, usually for free. trckr labels these "Open"..
How it scores
From independent testers, never the lab's own claims.
- Combined test score#17 of 239
- GPT-6.2 Astra174
- 2Claude Fable 5.2174
- 16Muse Wonder162
- 17Kimi K3.1 Pro161
- 18GPT-5.6 Terra161
Source: Epoch AI · Epoch Capabilities Index
- Chat answers, by people's votes#20 of 87
- Gemini 4 Carbon1,526
- 2Gemini 4 Argon1,521
- 19Gemini 3.1 Pro1,487
- 20Kimi K3.1 Pro1,487
- 21Gemini 3 Pro1,486
Source: LMArena · LMArena text, adjusted for style
- Building apps, by people's votes#15 of 87
- Claude Fable 5.21,824
- 2Claude Opus 5.51,818
- 14Claude Opus 51,691
- 15Kimi K3.1 Pro1,684
- 16GPT-6.1 Terra1,682
Source: LMArena · LMArena WebDev
Compare Kimi K3.1 Pro with
It came out 27 days after Nemotron 3.5 Super VL, the lab's previous model.
The details
NVIDIA's Quantisation: Shrinking a model by storing its numbers less precisely, so it fits on smaller computers. A little less accurate, much lighter. FP8, GGUF, AWQ and MLX are common formats. quantized copy of Moonshot AI's Kimi-K3.1-Pro, a large Multimodal: Works with more than text: a multimodal model can also look at images, listen to audio or watch video. model with 2.8T total and 104B Active parameters: In a mixture-of-experts model, the part that actually works on each word. A model with 74B parameters and 9B active needs the memory of a 74B model but runs about as fast as a 9B one. and a 1,048,576-token Context window: How much a model can keep in mind at once: your question, any files you share and its own replies. Measured in tokens; 1,000 tokens is about 750 words.. It accepts text and images, outputs text, and is free to download under NVIDIA's open model agreement, aimed at NVIDIA Blackwell GPU: The chip that runs AI models. Bigger models need more and larger GPUs, which is why the biggest ones live in data centres. systems.
Model ID: The exact name developers type to use a model, like gpt-5-mini. New IDs often show up in a lab's systems before the official announcement, which is how trckr spots many releases early.
nvidia/Kimi-K3.1-Pro-NVFP4Timeline
/Hugging FaceOfficialFirst signal
Get alerts like this the moment they happen.
From NVIDIA and every other major lab.