Provider guide
Where to run Llama 3.2 1B Instruct
3 live offers tracked — output prices vary 10.1× between the cheapest and most expensive host, so the provider choice matters as much as the model choice.
FASTEST MEASURED
153 tok/s
measured throughput · $0.027 input / $0.20 output per million tokens
Novita AICHEAPEST$0.020 in/1M$0.020 out/1M—131K—6h agoOpenRouter$0.027 in/1M$0.20 out/1M—60K—15m agoCloudflare Workers AI$0.027 in/1M$0.20 out/1M153 tok/s60K—15m ago
OpenRouterFull specs, hardware verdicts and benchmarks on theLlama 3.2 1B Instruct model page →