Updated

Kimi K3.1 Pro and K3.1 Pro Highspeed

Moonshot introduced Kimi K3.1 Pro, the first model in its K3.1 series, alongside K3.1 Pro Highspeed. The company describes a 3T-class open-weight model with a 1-million-token context window and improvements for coding, 3D creation, financial research, and agentic tasks.

By No Code MBA · Facts checked

Provider
Moonshot AI
Announcement date
Status at last check
Generally available
Model family
Kimi K3.1
Preceding model
Kimi K3
Capabilities
Text, Image, Video, Coding

What changed

Kimi K3.1 Pro is a post-trained version of Kimi K3 and the first model in the K3.1 series. Moonshot describes it as a 3T-class open-weight model with a 1,048,576-token context window and up to 1M output tokens, and says it improves coding, 3D creation, financial research and agentic tasks. It accepts text, images and uploaded video. A Highspeed variant runs 2 to 3 times faster than K3.1 Pro, according to Moonshot. Thinking is always on, with low, high and max reasoning effort settings.

Access and availability

Use the Kimi platform API with model IDs kimi-k3.1-pro or kimi-k3.1-pro-highspeed through the Chat Completions API. Access requires a top-up of at least $1, and rate limits scale with cumulative top-up. Moonshot says full open weights will be released by November 4.

Official access documentation

What you can build

Evaluate both variants on the same long-context coding or research task, comparing output quality, response time, and review effort on your own inputs.

Limitations

Open weights were not yet published at launch, and no license or exact parameter count was stated. Sampling parameters are fixed, thinking cannot be disabled, image input must be base64 or uploaded files (no public URLs), and built-in web search is not recommended for production. Moonshot says cyber training data was removed and offensive cyber capability is below frontier proprietary models.

Sources and release history