Kimi K3.1 Pro

Kimi K3.1 Pro

Moonshot's most capable open-weight model for long agentic work.

Moonshot's 2.8T-parameter open-weight flagship with a 1M-token context, built for long coding and agentic tasks.

Kimi K3.1 Pro benchmark results from Moonshot's tech blog

Kimi K3.1 Pro benchmark results from Moonshot's tech blog

Overview

What is Kimi K3.1 Pro?

Moonshot's most capable open-weight model for long agentic work.

Kimi K3.1 Pro is Moonshot AI's flagship model, released October 23, 2026. It is a heavily post-trained version of Kimi K3, keeping the same 2.8-trillion-parameter mixture-of-experts backbone with 104 billion parameters active per token, and a 1,048,576-token context window.

The model is built for agentic work: long coding sessions, 3D modeling, and finance tasks. Moonshot reports 65.7% on Terminal-Bench 4.0 in its own launch figures, just behind Claude Opus 5.5 at 66.4%, and says the model cuts Kimi K3's hard hallucination rate from 23.4% to 8.3%.

Unlike most flagships, Kimi K3.1 Pro is open-weight: the full weights were published on Hugging Face on November 4, 2026, as moonshotai/Kimi-K3.1-Pro (1.56 TB) under the Kimi K3.1-Pro License, so teams can self-host it. It is also available as an API at $2.00 per million input tokens and $10.00 per million output tokens, with a faster Highspeed variant at $4.00 and $20.00.

Platforms and languages

Kimi K3.1 Pro Availability

Platforms

WebApi

Languages

EnglishChinese

Capabilities

Kimi K3.1 Pro Key Features

Open weights

Full weights published on Hugging Face on November 4, 2026, under the Kimi K3.1-Pro License, so teams can download and self-host the model.

1M-token context

A 1,048,576-token window for whole repositories, long documents, and extended agent runs.

Agentic coding

Moonshot reports 65.7% on Terminal-Bench 4.0 in its own launch figures, close to Claude Opus 5.5 at 66.4%.

Highspeed variant

Kimi K3.1 Pro Highspeed runs 2 to 3 times faster than the standard model, at $4.00 input and $20.00 output per million tokens.

Lower hallucination rate

Moonshot says the claim-level hallucination rate on its hard internal test fell from 23.4% for Kimi K3 to 8.3% for K3.1 Pro.

Available everywhere Kimi runs

Live on kimi.com, the Kimi mobile apps, Kimi Work, Kimi Code, and the Kimi API under the model ID kimi-k3.1-pro.

Plans and access

Kimi K3.1 Pro Pricing

Paid

Common questions

Kimi K3.1 Pro FAQs

What is Kimi K3.1 Pro?

Kimi K3.1 Pro is Moonshot AI's flagship model, released October 23, 2026. It is a 2.8-trillion-parameter mixture-of-experts model with a 1-million-token context window, built for long coding and agentic tasks.

Is Kimi K3.1 Pro open source?

Its full weights were published on Hugging Face on November 4, 2026, as moonshotai/Kimi-K3.1-Pro under the Kimi K3.1-Pro License, which permits commercial use and serving with conditions for very large services.

How much does the Kimi K3.1 Pro API cost?

The standard model costs $2.00 per million input tokens and $10.00 per million output tokens, with cached input at $0.20. The faster Highspeed variant costs $4.00 and $20.00.

How does Kimi K3.1 Pro compare with Claude Opus 5.5?

Moonshot reports 65.7% on Terminal-Bench 4.0, just behind Claude Opus 5.5 at 66.4%. Independent evaluator Artificial Analysis scores Kimi at 46 on its Intelligence Index, against 59 for Opus 5.5.

Reviews

See what the community thinks and share your experience.

—

Based on 0 ratings

Rating distribution

0
0
0
0
0

Leave a review

Sign in to rate

Community reviews

No reviews

No written reviews yet.

Explore by category

Llms Foundation Models

View all Llms Foundation Models websites
Qwen4-Max preview
Qwen4-Max

Alibaba's flagship 2.4T-parameter model with a 1M-token context and native vision, priced from $1.50 per million input tokens.

Llms Foundation Models
Sarvam Prism preview
Sarvam Prism

Sarvam AI's open-weight System One decision model that scores fixed options in a single pass and returns calibrated probabilities for enterprise workflows.

Llms Foundation Models

Explore by category

AI Apis Sdks

View all AI Apis Sdks websites
Qwen4-Max preview
Qwen4-Max

Alibaba's flagship 2.4T-parameter model with a 1M-token context and native vision, priced from $1.50 per million input tokens.

AI Apis Sdks
FlashData preview
FlashData

YouTube Download, YouTube Data & Google Search APIs: deliver media to your cloud, retrieve existing captions and metadata, and get search JSON.

AI Apis Sdks
Inkling 2 Small preview
Inkling 2 Small

Open-weight multimodal reasoning model with 594B total parameters, a 1M-token context, and Apache 2.0 weights for self-hosting and fine-tuning.

AI Apis Sdks
CALL-E preview
CALL-E

Lets your AI assistant make real phone calls for you — you confirm, CALL-E dials, and you get the answer. 100 free credits for new users.

AI Apis Sdks
E2X preview
E2X

Prepaid AI video & image API, pay per request—up to 93% cheaper than rival platforms.

AI Apis Sdks