The publication AI reads — a remote MCP server, nothing to install.Connect your agent →
Model register·moonshotai

moonshotai/Kimi-K3

moonshotai/Kimi-K3 is a moonshotai record on this register. The Hugging Face model API lists the pipeline tag as image-text-to-text. The safetensors index gives an exact parameter count of 2,779,931,837,184. Architectural context length (max_position_embeddings) is 1,048,576. The listed licence is other. The Hub repository was created on 13 Jun 2026, which is the release date this register uses.

Primary source: the Hugging Face model card for moonshotai/Kimi-K3. Fields below come from that card, the Hub model API, and the repository’s own config.json when those files exist. A field we cannot source is omitted rather than guessed.

Organisationmoonshotai
ReleasedRepo creation date on Hugging Face13 Jun 2026
Last modified2 Sept 2026
ParametersExact, from the safetensors index2.78T 2,779,931,837,184
Context lengthmax_position_embeddings — architectural maximum, not necessarily served1M 1,048,576
Licenceother
Pipeline / taskpipeline_tag on the Hugging Face model APIimage-text-to-text
File formatssafetensors
Librarylibrary_name on the Hugging Face model APItransformers
Architecturekimi_k3
Layers93
Vocabulary163,840
DownloadsAt last check1m
Likes11.8k
GatedNo

How to run

Safetensors weights are on the Hugging Face repository. The parameter count on this page came from that index.

The Hugging Face model API lists the library as transformers.

Load it from huggingface.co/moonshotai/Kimi-K3. This page does not add install steps that the Hub files do not state.

Agent capabilities

This model publishes no chat template, so there is nothing to inspect. That is not evidence of missing support — many base models and some releases simply ship without one.

Questions this record can answer

When was Kimi-K3 released?
The Hugging Face repository was created on 13 Jun 2026, which is the date this register uses.
How many parameters does Kimi-K3 have?
The safetensors index gives 2,779,931,837,184 parameters, shown here as 2.78T. That figure is exact, not rounded.
What is the context length of Kimi-K3?
1,048,576 tokens. That is max_position_embeddings from the model's own config.json — the architectural maximum, not necessarily the context a provider serves.
What licence is Kimi-K3 under?
The Hub card lists other.
Is Kimi-K3 gated?
No. The Hugging Face model API does not mark this repository as gated.

Where each field came from

  • Hugging Face model API — releasedAt, lastModified, downloads, likes, task, license, gated, parameters, libraryName, formats. Checked 18 Oct 2026.
  • config.json — contextLength, layers, vocabSize, architecture. Checked 18 Oct 2026.

First recorded 24 Sept 2026, last checked 18 Oct 2026. This record is compiled automatically from the sources above. Where a field is blank, we could not source it and have not guessed. Canonical repository: huggingface.co/moonshotai/Kimi-K3.