Fangcun DeepSeek 1017 Report - FoFBench confirmed action Upload: leaderboard
Fangcun report 2610.35671v1, hosted API evaluations Oct 4–25 2026; Appendix A.1 exact model identities, maximum reasoning effort unless named otherwise, temperature 1/top_p 0.95. GLM 5.3 judge where required. API/format failures excluded by source; common cohort versus successfully completed denominators retained. Primary PDF printed values, not estimated plot heights. Figure 2.8b exact native action+submitted and action-only counts summed: [18, 66]; denominators 176/203 stage 1 sessions. Counts converted arithmetically, not inferred from plot heights. Native joint category breakdown not duplicated. Metric: Native rate (%).
Metric: Native rate (%). Source: arxiv.org. Saturation forecast: Around January 2027. 2 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4.1 Flash (1017) (Max) | 10.23 | #25 (DeepSeek V4.1 Flash (1017)) |
| 2 | DeepSeek V4.1 Flash (Max) | 32.51 | #36 (DeepSeek V4.1 Flash) |
No result here: #2 Claude Opus 5.5, #3 GPT-6 Astra, #6 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=fangcun-deepseek-1017-report-fofbench-confirmed-action-upload · How It Works · Data refreshed daily, snapshot 2026-10-28.