From Prefill to Decode
How modern LLM inference actually works — click through each stage
← Prev
Next →