DeepSeek Cascade Beats GPT-5.6 Luna on DeepSWE at 37% Lower Cost
togethercompute · x · 2026-08-08
Together Compute evaluated DeepSeek V4 Flash and GPT-5.6 Luna on the DeepSWE benchmark.
Results show that a DeepSeek-first cascade architecture, combined with test-suite verification, solved more coding tasks than using Luna alone, while simultaneously reducing the cost per task by 37%.
Related event: DeepSeek-V4 Flash Benchmarked: One-Sixth Cost, 80% of Luna's Performance(14 posts)→
More from coding & agent
- Constraints you force on models today will become tech debt in months, warns dev — zack_overflow · 2026-10-16
- Liquid AI tutorial: fine-tune your d1-3B decision model in 30 minutes on a free GPU — helloiamleonie · 2026-10-16
- Alibaba open-sources OpenGCUA, a full-stack generalist computer-use agent; +19.6 points on OSWorld-v2 across 8 models — Tongyi-MAI · 2026-10-16
- BotBeat: a social network where all users are AI agents, exposed via 32 MCP tools — Dismal_Piccolo4973 · 2026-10-16
- DKGI paper: agents write their own tools to search knowledge graphs, beating baselines by 6.7 Hit@1 — yuz9yuz · 2026-10-16
- Dev open-sources lgtmaybe: running PR code reviews on a local 27B model via hunk splitting — mattjcoles · 2026-10-16