Sonnet 5.5 (max) long-horizon agent runs: 1,833 turns, 6.5 hours, $192 with score dropping
PawelHuryn · x · 2026-09-30
PawelHuryn published a third long-horizon autonomous agent run of Sonnet 5.5 (max) across two repos:
- Run 1: score 57, 1,566 turns, 258 min, $154.05
- Run 2: score 54, 1,093 turns, 212 min, $115.55
- Run 3: score 43, 1,833 turns, 392 min, $192.52
Scores don't correlate with turns, time or cost — Run 3 burned the most turns, time and money while scoring lowest, highlighting high variance and unpredictable costs in long-horizon agentic tasks. A log-scale score-vs-turns view with model filters is available.
More from coding & agent
- Google Research open-sources PaperOrchestra, a multi-agent framework for writing AI papers — google-research · 2026-11-16
- ffmpeg-skill gives coding agents local video editing via 42 FFmpeg tools and MCP — kajisho5 · 2026-11-16
- Agents Now Reply to Slack Messages on Their Own, Resolving a Prod Incident Ghost-to-Ghost — iamrobotbear · 2026-11-16
- Engineer's Interview Prep Method: Use Claude to Find Which Answers Undersell Your AI Work — brandon_galang · 2026-11-16
- DeepSeek open-sources its app frontend "chat-demo", already at 3.9k stars — NielsRogge · 2026-11-16
- lemire: AI's real impact on coding is expanding the engineer's 'cognitive radius', not speed — lemire · 2026-11-16