library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen3-Coder-30B-A3B-Instruct/blob/main/LICENSE
pipeline_tag: text-generation
base_model: Qwen/Qwen3-Coder-30B-A3B-Instruct
tags:
- llama-cpp
- text-generation-inference
language:
- en
Qwen3-Coder-30B-A3B-Instruct-GGUF
Qwen3-Coder-30B-A3B-Instruct is a state-of-the-art large language model from the Qwen series, specifically optimized for advanced agentic coding, browser-based automation, and foundational programming tasks. Featuring 30.5 billion total parameters with 3.3 billion activated in a Mixture-of-Experts (MoE) architecture, it delivers strong performance and efficiency for complex code and tool-use scenarios. Its standout long-context capability natively processes up to 262,144 tokens—expandable to 1 million with Yarn—making it ideal for repository-scale code understanding and generation.
The model supports agentic coding with advanced function-call handling, and is compatible with popular local inference platforms like Ollama, LMStudio, and llama.cpp. Designed for both pretraining and post-training stages, Qwen3-Coder-30B-A3B-Instruct runs exclusively in non-thinking mode, ensuring fast, high-quality outputs for coding and automation workflows without requiring explicit configuration for thinking blocks
Use with llama.cpp
Install llama.cpp through brew (works on Mac and Linux)
brew install llama.cpp
Invoke the llama.cpp server or the CLI.
CLI:
llama-cli --hf-repo prithivMLmods/Qwen3-Coder-30B-A3B-Instruct-GGUF --hf-file qwen3-coder-30b-a3b-instruct-q4_k_m.gguf -p "The meaning to life and the universe is"
Server:
llama-server --hf-repo prithivMLmods/Qwen3-Coder-30B-A3B-Instruct-GGUF --hf-file qwen3-coder-30b-a3b-instruct-q4_k_m.gguf -c 2048
Note: You can also use this checkpoint directly through the usage steps listed in the Llama.cpp repo as well.
Step 1: Clone llama.cpp from GitHub.
git clone https://github.com/ggerganov/llama.cpp
Step 2: Move into the llama.cpp folder and build it with LLAMA_CURL=1 flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).
cd llama.cpp && LLAMA_CURL=1 make
Step 3: Run inference through the main binary.
./llama-cli --hf-repo prithivMLmods/Qwen3-Coder-30B-A3B-Instruct-GGUF --hf-file qwen3-coder-30b-a3b-instruct-q4_k_m.gguf -p "The meaning to life and the universe is"
or
./llama-server --hf-repo prithivMLmods/Qwen3-Coder-30B-A3B-Instruct-GGUF --hf-file qwen3-coder-30b-a3b-instruct-q4_k_m.gguf -c 2048