Instructions to use AesSedai/Kimi-K2.5-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AesSedai/Kimi-K2.5-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="AesSedai/Kimi-K2.5-GGUF",
	filename="IQ2_S/Kimi-K2.5-IQ2_S-00001-of-00008.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use AesSedai/Kimi-K2.5-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S
# Run inference directly in the terminal:
llama-cli -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S
# Run inference directly in the terminal:
llama-cli -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S
# Run inference directly in the terminal:
./llama-cli -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S
# Run inference directly in the terminal:
./build/bin/llama-cli -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S

Use Docker

docker model run hf.co/AesSedai/Kimi-K2.5-GGUF:IQ2_S

LM Studio
Jan
Ollama
How to use AesSedai/Kimi-K2.5-GGUF with Ollama:
```
ollama run hf.co/AesSedai/Kimi-K2.5-GGUF:IQ2_S
```

Unsloth Studio

How to use AesSedai/Kimi-K2.5-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for AesSedai/Kimi-K2.5-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for AesSedai/Kimi-K2.5-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for AesSedai/Kimi-K2.5-GGUF to start chatting

How to use AesSedai/Kimi-K2.5-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "AesSedai/Kimi-K2.5-GGUF:IQ2_S"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use AesSedai/Kimi-K2.5-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf AesSedai/Kimi-K2.5-GGUF:IQ2_S

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default AesSedai/Kimi-K2.5-GGUF:IQ2_S

Run Hermes

hermes

Docker Model Runner
How to use AesSedai/Kimi-K2.5-GGUF with Docker Model Runner:
```
docker model run hf.co/AesSedai/Kimi-K2.5-GGUF:IQ2_S
```

Lemonade

How to use AesSedai/Kimi-K2.5-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull AesSedai/Kimi-K2.5-GGUF:IQ2_S

Run and chat with the model

lemonade run user.Kimi-K2.5-GGUF-IQ2_S

List all available models

lemonade list

Updates

03/25/2026

I've re-quanted and uploaded new versions for the IQ2_XXS, IQ2_2, and IQ3_S quantizations. Those three are using a mixture of @eaddario's target-bpw PR along with some small changes I added to support --tensor-type overrides.

The result is that these quants perform better than my previous quants, and when I measured the old quants a couple of days ago it turns out there was some pretty catastrophic issues with the IQ3_S and the IQ2_S specifically. These new quants measure much better and should serve as better quality replacements.

I don't have specific FFN Up / Gate / Down mixtures for the IQ2_XXS, IQ2_S, and IQ3_S quants due to how the bpw budget selection works, but I've kept most of the model in high quality like the rest of my MoE-optimized quants.

02/11/2026

Vision support for K2.5 has been merged into llama.cpp's master branch and no longer needs to use the PR branch.

02/08/2026

I've updated the PR code to address feedback and updated the mmproj files here to be compatible with the new PR code.

02/01/2026

moonshotai has published an updated chat_template.jinja, I have updated the GGUFs in this repository so please re-download the first shard (00001) for your desired quant.

The default system prompt might cause confusion to users and unexpected behaviours, so we remove it.
The token <|media_start|> is incorrect; it has been replaced with <|media_begin|> in the chat template.

Model

This is a text-and-image-only GGUF quantization of moonshotai/Kimi-K2.5. This means that video input is not present in this GGUF, and will not be available until support is added upstream in llama.cpp.

MMPROJ files for image vision input have been provided, and support has been merged into the llama.cpp master branch recently.

This Q4_X quant is the "full quality" equivalent since the conditional experts are natively INT4 quantized directly from the original model, and the rest of the model is Q8_0. I also produced and tested a Q8_0 / Q4_K quant, the model size was identical and the PPL was barely higher. Their performance was about the same so I've only uploaded the Q4_X variant.

Quant	Size	Mixture	PPL	1-(Mean PPL(Q)/PPL(base))	KLD
Q4_X	543.62 GiB (4.55 BPW)	Q8_0 / Q4_0	1.8248 +/- 0.00699	0	0
IQ3_S	377.50 GiB (3.16 BPW)	Q8_0 / varies	2.116713 ± 0.008620	+16.0796%	0.158551 ± 0.001084
IQ2_S	311.71 GiB (2.61 BPW)	Q8_0 / varies	2.433594 ± 0.010455	+33.4572%	0.294937 ± 0.001721
IQ2_XXS	262.74 GiB (2.20 BPW)	Q8_0 / varies	3.119876 ± 0.014508	+71.0926%	0.540149 ± 0.002570