Garry Tan praises particle-1: perfect 15/15 benchmark score at 1/12th the cost of Opus 5.5
garrytan · x · 2026-10-30
Y Combinator president Garry Tan says he's been using particle-1 and is very impressed, sharing Gstack benchmark results:
- Perfect score: 15/15 passes across 5 gstack skill tasks × 3 runs, matching Claude Opus 5.5 and Claude Sonnet 5.5
- Cost edge: $0.007 per task — about 5× cheaper than Sonnet 5.5 ($0.037) and 12× cheaper than Opus 5.5 ($0.088), also below DeepSeek V4.1 Flash ($0.008)
- Method: automatic pass/fail checks (no AI judge), run in Claude Code via OpenRouter, preregistered; the whole benchmark cost $2.24
- Usage tip: great for everyday steps; keep your main model for final review. Try particle/particle-1 in Capy or particle-ai/particle-1 on OpenRouter
Related event: Particle-1 Gains Garry Tan's Endorsement, Lands on OpenRouter and Vercel(3 posts)→
More from coding & agent
- Claude API docs reveal beta Agents, Sessions and Environments APIs with managed sandboxes — aigclink · 2026-10-30
- Graph embeddings in one line: FastRP by default, GraphSAGE for attributes — JeremyCMorgan · 2026-10-30
- CodeRabbit CEO: $8M seed rounds no longer enough as 'software factories' need tokens; Garry Tan pushes back — garrytan · 2026-10-30
- GBrain Cloud launches a CLI: one command spins up a cloud AI workspace — garrytan · 2026-10-30
- Designing a multi-bot AI automation system to run one person's life and three businesses — Xeriphy · 2026-10-30
- "The agents of the internet of things" — a catchy frame for device-side agents — sull · 2026-10-30