Instructions to use prithivMLmods/Ripple-65B-Preview-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use prithivMLmods/Ripple-65B-Preview-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="prithivMLmods/Ripple-65B-Preview-GGUF")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("prithivMLmods/Ripple-65B-Preview-GGUF", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use prithivMLmods/Ripple-65B-Preview-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 prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf prithivMLmods/Ripple-65B-Preview-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 prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M

Use Docker

docker model run hf.co/prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M

Llama new
LM Studio
Jan

vLLM

How to use prithivMLmods/Ripple-65B-Preview-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "prithivMLmods/Ripple-65B-Preview-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": "prithivMLmods/Ripple-65B-Preview-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M

SGLang

How to use prithivMLmods/Ripple-65B-Preview-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "prithivMLmods/Ripple-65B-Preview-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "prithivMLmods/Ripple-65B-Preview-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "prithivMLmods/Ripple-65B-Preview-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "prithivMLmods/Ripple-65B-Preview-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use prithivMLmods/Ripple-65B-Preview-GGUF with Ollama:
```
ollama run hf.co/prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M
```
Unsloth Desktop

How to use prithivMLmods/Ripple-65B-Preview-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use prithivMLmods/Ripple-65B-Preview-GGUF with Docker Model Runner:
```
docker model run hf.co/prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M
```

Lemonade

How to use prithivMLmods/Ripple-65B-Preview-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Ripple-65B-Preview-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use prithivMLmods/Ripple-65B-Preview-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 prithivMLmods/Ripple-65B-Preview-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 prithivMLmods/Ripple-65B-Preview-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use prithivMLmods/Ripple-65B-Preview-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prithivMLmods/Ripple-65B-Preview-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 "prithivMLmods/Ripple-65B-Preview-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"

Ripple-65B-Preview-GGUF

Ripple-65B Preview is Ricursive Intelligence's 65B dense Transformer, pretrained from scratch on 12T tokens of mostly English text and code, with up to 1M context. This repo holds the BF16 instruction-tuned model, and NVFP4, base and DFlash2 versions are released separately. Post-training was light and focused on agentic coding. The card reports 80.5 on Terminal-Bench 2.1 and 79.1 on SWE-bench Pro v2.0 Hard, the best in its table, ahead of much larger MoE models such as Nemotron 3 Ultra 550B, Inkling 975B, Muse Spark Open 816B and Reflection Beam 501B. It trails them on knowledge (GPQA Diamond 85.2) and on some agentic suites such as SWE-bench Verified (76.5) and MCP Atlas (71.67). Some comparison scores were measured in-house, and the card says agentic results are hard to reproduce. The NVFP4 build is meant to fit consumer hardware such as the DGX Spark, and the model is released under Apache 2.0.

ripple-65b preview from ricursive intelligence now supports mainline llama.cpp, as pull request #31692 (model: add ripple (ricursiveai)) has been merged into llama.cpp.

Model Files

File Name	Quant Type	File Size	File Link	Description
Ripple-65B-Preview.BF16.gguf	BF16	130 GB	Link	Full BF16 weights. Highest quality, largest file size.
Ripple-65B-Preview.Q3_K_L.gguf	Q3_K_L	33.6 GB	Link	Lower quality but usable, good for low RAM availability.
Ripple-65B-Preview.Q4_K_M.gguf	Q4_K_M	38.8 GB	Link	Good quality, default size for most use cases, recommended.
Ripple-65B-Preview.Q5_K_M.gguf	Q5_K_M	45.8 GB	Link	High quality, recommended.
Ripple-65B-Preview.Q8_0.gguf	Q8_0	68.8 GB	Link	Extremely high quality, generally unneeded but max available standard quant.

Q4_K_M Test Run

Tested Ripple-65B-Preview.Q4_K_M.gguf on 2× NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs (96 GB VRAM each, 192 GB total), using llama-cli with full GPU offloading.

Performance Metrics

Metric	Value
Average Prompt Processing Speed	390.22 tokens/s
Average Generation Speed	33.04 tokens/s
Total Wall Time	106.9 seconds
Context Length	16,384 tokens
KV Cache (K)	`q8_0`
GPU	2× NVIDIA RTX PRO 6000 Blackwell Server Edition
GPU VRAM Capacity	192 GB (96 GB × 2)
Peak GPU VRAM Usage	39.05 GiB (19.78 GiB GPU 0 + 19.27 GiB GPU 1)
Peak GPU Utilization	55% (GPU 0) / 54% (GPU 1)
GPU Layers Offloaded	Auto (`-ngl auto`)
Random Seed	42
Speculative Decoding	No
Test Runs	5

Test Shots

Shot	Test
Shot 1	C: Insert Node at End of Singly Linked List
Shot 2	Jogger and Train Passing Time
Shot 3	Two Pipes Filling a Tank
Shot 4	Students in Examination Rooms
Shot 5	Father and Son Ages

Run Inference

./build/bin/llama-cli \
  -m /path/to/models/Ripple-65B-Preview.Q4_K_M.gguf \
  -ngl auto \
  -c 16384 \
  --cache-type-k q8_0 \
  --seed 42 \
  -n 8192 \
  -st \
  -p "Your prompt here"

Performance results reflect this specific hardware and inference configuration. Actual throughput may vary depending on the prompt, model, software build, and runtime settings.