// topics
1 posts, newest first.
← Blog · Search · Archive
29 Oct 2026 · ia, custo, llm · 5 min
Price table for five options in October 2026 plus one measurement of my own: the qwen2.5:0.5b model on CPU generated 23.8 tokens per second, or 11.7 hours for one million tokens. Local cost is hardware, not the token.
You reached the end. Every post is here.