Qwen2.5-0.5B-Instruct
Alibaba's half-billion-parameter instruction model for edge and mobile.
About
Qwen2.5-0.5B-Instruct is the instruction-tuned 0.5B variant of the Qwen2.5 family, released by Alibaba's Qwen team in September 2024. Despite its tiny size it retains usable instruction following and multilingual ability across 29 plus languages, with a 32K token context window that is unusual for a model this small.
It is a favorite for on-device and browser-adjacent experimentation because the full weights are only about 1GB. The model is released under Apache 2.0, making it free for research and commercial use, and it has spawned hundreds of community fine-tunes.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 124; Trending score 76/100 from researcher assessment.