Moonshot AI

Kimi-K2-Instruct-0905

Generally availableOpen weights262K context

Summary

Kimi-K2-Instruct-0905 is an open-weight language model from Moonshot AI, released on 5 Sept 2025. It accepts text and generates text, with a 262K-token context window and up to 16K output tokens per response. As of 18 Nov 2026, the lowest listed API price is $0.60 per million input tokens and $2.50 per million output tokens (Novita AI), across 3 providers we track.

Kimi model for long-context chat, coding, and agentic reasoning

Last verified 18 Nov 2026. Every figure on this page links to its source.

Specifications

DeveloperMoonshot AI
Release date5 Sept 2025 [models.dev]
StatusGenerally available [models.dev]
WeightsOpen weights [Hugging Face]
Licensemodified-mit [Hugging Face]
Parameters1T [Hugging Face]
Context window262K (262,144 tokens) [models.dev]
Max output16K tokens [models.dev]
Knowledge cutoffOct 2024 [models.dev]
InputText [models.dev]
OutputText [models.dev]
Reasoning modeNo [models.dev]
Tool callingYes [models.dev]
Structured outputYes [models.dev]
Hugging Facemoonshotai/Kimi-K2-Instruct-0905 [Hugging Face]

Kimi-K2-Instruct-0905 API pricing by provider

USD per million tokens. Sorted by input price.

ProviderInputOutputCached inputContextSource
Novita AI
moonshotai/kimi-k2-0905
$0.60$2.50–262Kmodels.dev, 18 Nov 2026
OpenRouter
moonshotai/kimi-k2-0905
$0.60$2.50–262KOpenRouter, 18 Nov 2026
Hugging Face Inference Providers
moonshotai/Kimi-K2-Instruct-0905
$1$3–262Kmodels.dev, 18 Nov 2026

Running Kimi-K2-Instruct-0905 locally

Estimated memory, from the parameter count and the architecture in config.json on Hugging Face. Includes about 0.8 GB of runtime overhead.

QuantizationWeightsTotal, 8K contextTotal, 32K context
Q4_K_M (4-bit)622 GB624 GB625 GB
Q8_0 (8-bit)1091 GB1092 GB1094 GB
FP16 / BF16 (16-bit)2053 GB2054 GB2056 GB

At 4-bit with an 8K conversation, Kimi-K2-Instruct-0905 needs about 624 GB, more than any single consumer graphics card; it needs a unified-memory machine, a multi-GPU setup or a split between graphics card and system RAM. Check your own hardware, or see Best computer for local AI on our sister site AIPCs.

Frequently asked questions

What is the context window of Kimi-K2-Instruct-0905?

Kimi-K2-Instruct-0905 has a context window of 262,144 tokens (262K) and can generate up to 16,384 tokens in a single response, according to models.dev.

How much does the Kimi-K2-Instruct-0905 API cost?

Kimi-K2-Instruct-0905 is listed from $0.60 per million input tokens and $2.50 per million output tokens (Novita AI). Prices last verified 18 Nov 2026.

Is Kimi-K2-Instruct-0905 open source?

Kimi-K2-Instruct-0905 is an open-weight model: its weights are published on Hugging Face as moonshotai/Kimi-K2-Instruct-0905 under the modified-mit license. Check the license terms before commercial use.

When was Kimi-K2-Instruct-0905 released?

Moonshot AI released Kimi-K2-Instruct-0905 on 5 Sept 2025, according to models.dev.

Which providers offer Kimi-K2-Instruct-0905?

We track Kimi-K2-Instruct-0905 on 3 providers: Novita AI, OpenRouter, Hugging Face Inference Providers.

How much VRAM does Kimi-K2-Instruct-0905 need?

About 624 GB at 4-bit (Q4_K_M) with an 8K-token conversation, and about 1092 GB at 8-bit. That is more than any single consumer graphics card holds. These are estimates from the model's published parameter count and architecture.

Similar models from other labs

Comparable on output type, weights, release date, price and context window. Prices are the lowest listed, USD per million tokens.

ModelLabReleasedContextInput $/1MOutput $/1MWeights
Qwen3-235B-A22B-Thinking-2507Alibaba (Qwen)25 Jul 2025262K$0.30$3Open
GLM-4.6Zhipu AI (Z.ai)30 Sept 2025205K$0.50$2Open
Hermes-4-405BNousresearch26 Aug 2025131K$1$3Open
GLM-4.5Zhipu AI (Z.ai)28 Jul 2025131K$0.60$2.20Open
Qwen3.5 397B-A17BAlibaba (Qwen)15 Feb 2026262K$0.45$3Open
DeepSeek-R1-0528DeepSeek28 May 2025164K$0.50$2.15Open

Change history

  • Now listed on Hugging Face Inference Providers
  • Knowledge cutoff changed from Dec 2024 to Oct 2024
  • Max output changed from 98K to 16K
  • Name changed from Kimi K2 Instruct 0905 to Kimi-K2-Instruct-0905
  • Parameter count changed from 1T to 1T
  • Knowledge cutoff changed from 1 Jan 2025 to Dec 2024
  • Max output changed from 64K to 98K
  • Name changed from Kimi-K2-0905 to Kimi K2 Instruct 0905
  • Context window changed from 256K to 262K
  • Knowledge cutoff changed from Dec 2024 to 1 Jan 2025
  • Name changed from Kimi K2 0905 to Kimi-K2-0905
  • Now listed on Novita AI
  • Knowledge cutoff changed from 31 Dec 2024 to Dec 2024
  • Release date changed from 4 Sept 2025 to 5 Sept 2025
  • Max output changed from 98K to 64K
  • Context window changed from 262K to 256K

Sources