Mellum 2.1 Thinking

by JetBrains

A model that thinks, on your own PC.

↓ 17.6 k downl. Apache-2.0 FreeLocalOpen weights

View on Hugging Face →

What it does

12 billion parameters, only 2.5 active: fast
Visible reasoning and a 131,000-token context
4-bit version: 8 GB, fits a 12 GB graphics card

Install it

Requirements: About 8 GB of memory for the 4-bit version (Q4_K_M). Ollama, LM Studio or llama.cpp.

With Ollama

ollama run hf.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF:Q4_K_M

With llama.cpp (OpenAI-compatible API)

llama-server -hf JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF:Q4_K_M --ctx-size 131072 --temp 0.6 --top-p 0.95 --top-k 20

The prompt for your agent

Paste it into your coding agent: it checks the requirements and installs it for you.

Claude CodeCodexCursorGemini CLIopencode

Prompt

Install JetBrains' Mellum 2.1 Thinking model locally.
1. Check how much memory is available (GPU and RAM): the Q4_K_M version needs about 8 GB.
2. Check that Ollama is installed; if it isn't, ask me before installing it.
3. Run: ollama run hf.co/JetBrains/Mellum2.1-12B-A2.5B-Thinking-GGUF:Q4_K_M
4. Ask it a test question and show me its answer, reasoning included (<think> tags).

Always review what your agent is about to run, especially install commands.