Qwen 2.5 0.5B

mobile-tiny

Ultra-compact model for low-RAM devices. Handles extraction and classification.

Parameters0.5B
Context32,768 tokens
Download0.4 GB
Min RAM3 GB
Architectureqwen2

Available quantizations

QuantSizeQuality
Q4_K_M0.4 GBgood
Q8_00.6 GBbest

Check whether your machine can run this model — with real measured speeds from anonymous community benchmarks — on the Central-Intel compatibility page.