5 yesterday

12B — GGUF — sampling | ~8-10 GB

tools thinking
ollama run decadence/mellum2.1:12b-a2.5b-thinking-q4_k_m

Applications

Claude Code
Claude Code ollama launch claude --model decadence/mellum2.1:12b-a2.5b-thinking-q4_k_m
OpenCode
OpenCode ollama launch opencode --model decadence/mellum2.1:12b-a2.5b-thinking-q4_k_m
Hermes Agent
Hermes Agent ollama launch hermes --model decadence/mellum2.1:12b-a2.5b-thinking-q4_k_m
OpenClaw
OpenClaw ollama launch openclaw --model decadence/mellum2.1:12b-a2.5b-thinking-q4_k_m

Models

View all →

Readme

Mellum2.1-12B-A2.5B-Thinking

model type quant size source
mellum2.1:12b-a2.5b-thinking-q4_k_m GGUF Q4_K_M 8.1 GB JetBrains
mellum2.1:12b-a2.5b-thinking-q6_k GGUF Q6_K 10 GB JetBrains

Sampling Parameters

Following Mellum2.1 recommended sampling parameters:

parameter value
temperature 0.6
top_p 0.95
top_k 20