What are the hardware requirements for Kimi K2 Instruct 0905?

Kimi K2 Instruct 0905 requires approximately 667.4 GB of VRAM and supports a context window of 256K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Kimi K2 Instruct 0905 and how large is it?

Kimi K2 Instruct 0905 is developed by unsloth, a model with 1026.5b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Kimi K2 Instruct 0905?

Kimi K2 Instruct 0905 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Kimi K2 Instruct 0905 by unsloth — VRAM 667.4GB, 256K context

Name: Kimi K2 Instruct 0905
Author: unsloth

Kimi K2 Instruct 0905 is an open-source language model by unsloth. Features: 1026.5b LLM, VRAM: 667.4GB, Context: 256K, License: other, Instruction-Based, LLM Explorer Score: 0.2.

Base model:moonshotai/kimi-k2-... Base model:quantized:moonshota... Conversational Custom code Deepseek v3 Endpoints compatible Fp8 Instruct Region:us Safetensors Sharded Tensorflow Unsloth

Model Card on HF 🤗: https://huggingface.co/unsloth/Kimi-K2-Instruct-0905

Kimi K2 Instruct 0905 Parameters and Internals

LLM Name	Kimi K2 Instruct 0905
Repository 🤗	https://huggingface.co/unsloth/Kimi-K2-Instruct-0905
Base Model(s)	Kimi K2 Instruct 0905 moonshotai/Kimi-K2-Instruct-0905
Model Size	1026.5b
Required VRAM	667.4 GB
Updated	2026-07-15
Maintainer	unsloth
Model Type	deepseek_v3
Instruction-Based	Yes
Model Files	0.5 GB: 1-of-62 17.1 GB: 2-of-62 17.1 GB: 3-of-62 17.1 GB: 4-of-62 17.1 GB: 5-of-62 17.1 GB: 6-of-62 17.1 GB: 7-of-62 17.1 GB: 8-of-62 17.1 GB: 9-of-62 17.1 GB: 10-of-62 17.1 GB: 11-of-62 17.1 GB: 12-of-62 17.1 GB: 13-of-62 17.1 GB: 14-of-62 17.1 GB: 15-of-62 17.1 GB: 16-of-62 17.1 GB: 17-of-62 17.1 GB: 18-of-62 17.1 GB: 19-of-62 17.1 GB: 20-of-62 17.1 GB: 21-of-62 17.1 GB: 22-of-62 17.1 GB: 23-of-62 17.1 GB: 24-of-62 17.1 GB: 25-of-62 17.1 GB: 26-of-62 17.1 GB: 27-of-62 17.1 GB: 28-of-62 17.1 GB: 29-of-62 17.1 GB: 30-of-62 17.1 GB: 31-of-62 17.1 GB: 32-of-62 17.1 GB: 33-of-62 17.1 GB: 34-of-62 17.1 GB: 35-of-62 17.1 GB: 36-of-62 17.1 GB: 37-of-62 17.1 GB: 38-of-62 17.1 GB: 39-of-62 17.1 GB: 40-of-62
Model Architecture	DeepseekV3ForCausalLM
License	other
Context Length	262144
Model Max Length	262144
Transformers Version	4.56.1
Vocabulary Size	163840

Rank the Kimi K2 Instruct 0905 capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Kimi K2 Instruct 0905

Best Alternatives	Context / RAM	Downloads	Likes
Kimi K2 Instruct	128K / 652.7 GB	33	22
Kimi K2 Instruct GGUF	128K / GB	47901	229

Note: green Score (e.g. "73.2") means that the model is better than unsloth/Kimi-K2-Instruct-0905.