Instructions to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

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 Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

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 Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

Use Docker

docker model run hf.co/Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Radamanthys11/Kimi-K2.5-DFlash-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Radamanthys11/Kimi-K2.5-DFlash-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

Ollama
How to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with Ollama:
```
ollama run hf.co/Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M
```

Unsloth Studio

How to use Radamanthys11/Kimi-K2.5-DFlash-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 Radamanthys11/Kimi-K2.5-DFlash-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 Radamanthys11/Kimi-K2.5-DFlash-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Radamanthys11/Kimi-K2.5-DFlash-GGUF to start chatting

How to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

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": "Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

OpenClaw new

How to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Docker Model Runner
How to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with Docker Model Runner:
```
docker model run hf.co/Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M
```

Lemonade

How to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Kimi-K2.5-DFlash-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use Radamanthys11/Kimi-K2.5-DFlash-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

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 Radamanthys11/Kimi-K2.5-DFlash-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

Kimi-K2.5-DFlash-GGUF

File size: 1,237 Bytes

519b4bd
b67e536
519b4bd
 
 
 
 
 
 
 
 
 
 
 
b67e536
 
519b4bd
 
105b1eb
519b4bd
 
 
105b1eb
 
 
 
 
519b4bd

---
base_model: z-lab/Kimi-K2.5-DFlash
tags:
- gguf
- dflash
- ik_llama.cpp
library_name: gguf
pipeline_tag: text-generation
---

# Kimi K2.5 DFlash GGUF

GGUF made to use in [ikawrakow/ik_llama.cpp](https://github.com/ikawrakow/ik_llama.cpp), currently for PR [#1970](https://github.com/ikawrakow/ik_llama.cpp/pull/1970). The small quantizations delivered here are made for test purposes; feel free to create your own quantization.

Derived from the safetensors DFlash draft model [z-lab/Kimi-K2.5-DFlash](https://huggingface.co/z-lab/Kimi-K2.5-DFlash).

## Compatible target model

- `Kimi-K2.5.gguf` - Mainly tested with IQ1_KT.

## Files

| File | Quant | Size |
|---|---|---|
| `kimi-k2-5-dflash-F16.gguf` | F16 | 6.5 GB |
| `kimi-k2-5-dflash-Q8_0.gguf` | Q8_0 | 3.5 GB |
| `kimi-k2-5-dflash-Q4_K_M.gguf` | Q4_K_M | 1.9 GB |

## Usage

```bash
./build/bin/llama-server \
  --model <target.gguf> \
  --model-draft <draft.gguf> \
  --spec-type dflash:n_max=<N>,cross_ctx=<N> ...
```

## Notes

- This repo contains DFlash draft models, not a standalone instruct model.
- Use it with the matching target family listed above.
- `Q4_K_M` and `Q8_0` are small test-oriented quants; create your own quant if you need a different tradeoff.