Local 27B ternary-compressed model builds an entire game in 5 hours on an RTX 3060 12GB, zero hand-written code
mishig25 · x · 2026-09-24
Developer @sudoingX shows a full game built entirely by a local AI model running on an RTX 3060 12GB — the #1 GPU on Steam.
- Model: PrismML's bonsai 2 27B (Qwen 3.8 27B dense compressed to ternary), only 5.95GB of weights, with MTP.
- Workflow: hermes agent ran for 5 hours, wrote 328k tokens across 8 JS files / 2,368 lines — zero hand-written code.
- Performance: 50 tok/s fresh, 22 tok/s average, 125k context with strong thread retention; suited for overnight batch work.
- A 12-minute video condenses the full 5-hour run plus gameplay.
More from coding & agent
- KAIST's WikiTrace: evidence-linked memory fixes agent "memory-induced degradation" — dair_ai · 2026-11-11
- Why inference speed will be AI products' biggest competitive edge in 1–2 years — itsOmSarraf_ · 2026-11-11
- What happened when I moved vector retrieval into Snowflake SQL instead of orchestrating RAG outside — dima806_dima · 2026-11-11
- OpenAI engineer runs Codex for 26 days straight hunting Atlantis, spawning "vibe archaeology" — Polymarket · 2026-11-11
- He replaced a $5,000/month ghostwriter with 11 Claude Skills and earned $47,000 from one post — aryanXmahajan · 2026-11-11
- Refero Styles ships 2,000+ DESIGN.md files to give Claude Code real design taste — bbssppllvv · 2026-11-11