How to use from
OpenClaw
Start the llama.cpp server
# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Mast-LLM/Gargantua-Code-Gen2: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://127.0.0.1:9931/v1 \
  --custom-model-id "Mast-LLM/Gargantua-Code-Gen2: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"
Quick Links

Gargantua Code Gen 2

Qwen2.5-Coder 7B Instruct fine-tuned (LoRA) for the openTARS app: it writes, reads, fixes and explains code using the app's prompt and its 19 tools. Quantized to Q4_K_M (4.7 GB). Text only, no vision.

Ollama: ollama run ProjectOpenTARS/Gargantua-Code-Gen2:code

Measured (04/11/2026, openTARS 5.2, 60 code tasks, one run)

51/60 (85%) on tasks that cover fixing code, explaining code, writing functions and scripts, and the shell. No dangerous actions. Average 7.7 s per task on an RTX 3050. The tasks are simple, so the score is not a general programming benchmark.

Source and tests: https://github.com/MastLLM/OpenTARS (MIT for the app code).


Made by MastLLM, makers of openTARS. Other labs keep building bigger. We build small.

Downloads last month
104
GGUF
Model size
8B params
Architecture
qwen2
Hardware compatibility
Log In to add your hardware

4-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for Mast-LLM/Gargantua-Code-Gen2

Base model

Qwen/Qwen2.5-7B
Quantized
(288)
this model