What are the hardware requirements for Kimi K2 Instruct BF16?

Kimi K2 Instruct BF16 requires approximately 1367.3 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Kimi K2 Instruct BF16 and how large is it?

Kimi K2 Instruct BF16 is developed by unsloth, a model with 1026.4b 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 BF16?

Kimi K2 Instruct BF16 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Kimi K2 Instruct BF16 by unsloth — VRAM 1367.3GB, 128K context

Name: Kimi K2 Instruct BF16
Author: unsloth

Kimi K2 Instruct BF16 is an open-source language model by unsloth. Features: 1026.4b LLM, VRAM: 1367.3GB, Context: 128K, License: other, Instruction-Based, LLM Explorer Score: 0.19.

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

Model Card on HF 🤗: https://huggingface.co/unsloth/Kimi-K2-Instruct-BF16

Kimi K2 Instruct BF16 Parameters and Internals

LLM Name	Kimi K2 Instruct BF16
Repository 🤗	https://huggingface.co/unsloth/Kimi-K2-Instruct-BF16
Base Model(s)	Kimi K2 Instruct moonshotai/Kimi-K2-Instruct
Model Size	1026.4b
Required VRAM	1367.3 GB
Updated	2026-07-28
Maintainer	unsloth
Model Type	deepseek_v3
Instruction-Based	Yes
Model Files	3.3 GB: 1-of-61 34.1 GB: 10-of-61 34.1 GB: 11-of-61 34.1 GB: 12-of-61 34.1 GB: 13-of-61 34.1 GB: 14-of-61 34.1 GB: 15-of-61 34.1 GB: 16-of-61 34.1 GB: 17-of-61 34.1 GB: 18-of-61 34.1 GB: 19-of-61 34.1 GB: 2-of-61 34.1 GB: 20-of-61 34.1 GB: 21-of-61 34.1 GB: 22-of-61 34.1 GB: 23-of-61 34.1 GB: 24-of-61 34.1 GB: 25-of-61 34.1 GB: 26-of-61 34.1 GB: 27-of-61 34.1 GB: 28-of-61 34.1 GB: 29-of-61 34.1 GB: 3-of-61 34.1 GB: 30-of-61 34.1 GB: 31-of-61 34.1 GB: 32-of-61 34.1 GB: 33-of-61 34.1 GB: 34-of-61 34.1 GB: 35-of-61 34.1 GB: 36-of-61 34.1 GB: 37-of-61 34.1 GB: 38-of-61 34.1 GB: 39-of-61 34.1 GB: 4-of-61 34.1 GB: 40-of-61 34.1 GB: 41-of-61 34.1 GB: 42-of-61 34.1 GB: 43-of-61 34.1 GB: 44-of-61 34.1 GB: 45-of-61 34.1 GB: 46-of-61
Model Architecture	DeepseekV3ForCausalLM
License	other
Context Length	131072
Model Max Length	131072
Transformers Version	4.53.2
Vocabulary Size	163840
Torch Data Type	bfloat16

Rank the Kimi K2 Instruct BF16 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 BF16

Best Alternatives	Context / RAM	Downloads	Likes
...i K2 Instruct 0905 Mlx DQ3 K M	256K / 204.6 GB	2234	12
Kimi K2 Instruct 0905 BF16	256K / 1399.1 GB	31	4
Kimi K2 Instruct 4bit	128K / 143.2 GB	5265	13

Note: green Score (e.g. "73.2") means that the model is better than unsloth/Kimi-K2-Instruct-BF16.