d3 is a new Ollama model that Makes DECISIONS (500 Messages in 23 Seconds!) - d3 on a Laptop GPU

d3 is a new Ollama model that never writes a sentence: it scores your choices. I ran it against Qwen3-4B on 500 labelled messages and 30 screenshots on an RTX 4060 8 GB. Chapters: 0:00 The result: 500 messages in 23 seconds 0:20 The pain: why chat models are awkward for sorting 0:41 What d3 is (Ollama page) 1:04 First run: CLI and API 1:25 The 500-message test 1:47 Results: d3 vs Qwen3-4B 2:15 The real difference: confidence scores 2:35 Screenshot test 3:01 Verdict Results (my run, Oct 26 2026): d3:2b 95.8% / 23.4 msg/s; d3:4b 98.4% / 6.2 msg/s; Qwen3-4B instruct 95.6% / 19.4 msg/s. d3:2b answers with confidence at or above 0.9: 406 of 500, all correct. Test data is synthetic (template-generated), so real tickets will be harder. Links: https://ollama.com/library/d3 | https://github.com/ollama/ollama/releases/tag/v0.40.11 If you are looking for cheap and reliable GPUs, I use RunPod: https://get.runpod.io/pe48 Get GPUs Runpod: https://get.runpod.io/pe48 Get Hostinger: https://www.hostg.xyz/SHJiu Coupon Code "PROMPT" 10% discount 🔗 Connect with me: Email → [email protected] Ko-fi → https://ko-fi.com/promptengineer Patreon → https://www.patreon.com/PromptEngineer975 Book a Call → https://calendly.com/prompt-engineer48/call GitHub → https://github.com/PromptEngineer48 Twitter/X → https://twitter.com/prompt48 Website → https://promptengineer48.com/