When was Kimi K2-Instruct-0905 released?

Kimi K2-Instruct-0905 was released on September 5, 2025 by MoonshotAI. This is the official Kimi K2-Instruct-0905 release date tracked on LLM Stats.

How big is Kimi K2-Instruct-0905?

Kimi K2-Instruct-0905 has 1000 billion parameters. It was trained on 15.5 trillion tokens. It ships as an open-weight model, so you can download and run it on your own hardware.

Who created Kimi K2-Instruct-0905?

Kimi K2-Instruct-0905 was created by MoonshotAI.

What is the license for Kimi K2-Instruct-0905?

Kimi K2-Instruct-0905 is released under the MIT license. This is an open-source / open-weight license that permits self-hosting.

Where is the Kimi K2-Instruct-0905 paper or technical report?

Kimi K2-Instruct-0905 has a paper or technical report available at https://github.com/MoonshotAI/Kimi-K2/blob/main/tech_report.pdf. Use that source for architecture, training, release and evaluation details.

What models should I compare Kimi K2-Instruct-0905 against?

Common Kimi K2-Instruct-0905 comparisons include Kimi K2-Instruct-0905 vs GPT OSS 120B High, Kimi K2-Instruct-0905 vs Nova 2 Lite, Kimi K2-Instruct-0905 vs Kimi K2 Instruct. Compare them side by side for benchmark scores, pricing, context window, latency and provider availability.

Kimi K2-Instruct-0905 Benchmarks, Pricing & Context Window

Name: Kimi K2-Instruct-0905
Author: MoonshotAI

Kimi K2-Instruct-0905: Benchmarks, Pricing & Context Window

Kimi K2-Instruct-0905 is a language model from MoonshotAI, released in September 2025.

Kimi K2-Instruct-0905 is the latest, most capable version of Kimi K2, achieving state-of-the-art performance in frontier knowledge, math, and coding among non-thinking models. This Mixture-of-Experts model features 32 billion activated

Kimi K2-Instruct-0905 benchmarks

Capability tiers

Standing within each category, adjusted for leaderboard depth.

Real tasks performance

High-confidence performance for Kimi K2-Instruct-0905 across real-world prompt categories. Only 95% intervals at most 4 points wide are shown.

Performance by conversation depth

How Kimi K2-Instruct-0905 holds up as conversations get longer.

Quality Tracker

Kimi K2-Instruct-0905 Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

llm-stats.com - Thu Oct 08 2026

Notice missing or incorrect data?

Kimi K2-Instruct-0905 model size

Kimi K2-Instruct-0905 has 1 trillion parameters and was trained on 15.5 trillion tokens. See how it compares to other models in the same parameter range.

ParametersTraining tokens

1TMoE

15.5Ttokens

16× tokens-to-params ratio

Frontier (200B+)

1B7B70B405B

Try now

huggle

Kimi K2-Instruct-0905in Huggle

Make it with
Kimi K2-Instruct-0905.

Kimi K2-Instruct-0905 latency

Kimi K2-Instruct-0905 time to first token, sustained output throughput, and failed-request rate from live model usage over the trailing 7 days.

Kimi K2-Instruct-0905 examples

Recent arena outputs from Kimi K2-Instruct-0905, picked from the highest-ranked matchups.

Kimi K2-Instruct-0905 license

Kimi K2-Instruct-0905 is released under the MIT license, which permits commercial use, has 1.0T parameters.

License: MIT; Commercial use allowed
Parameters: 1.0T

MIT License - allows commercial use

Kimi K2-Instruct-0905 resources

Official sources for Kimi K2-Instruct-0905: provider documentation, official playground, paper or system card, official launch post, source repository, model weights.

Kimi K2-Instruct-0905 vs other models

The most-compared alternatives to Kimi K2-Instruct-0905 are GPT OSS 120B High, Nova 2 Lite, Kimi K2 Instruct. Open any pair side-by-side for benchmarks, pricing, context, and latency.