November 2026
Does the standard hold on each new generation?
A generation is compared only with its predecessor in the same run: same requests, same judges, same standard.
Each change gets one of three verdicts, and it counts only when its interval clears zero and it's at least
5 points with guidance or 10 points without.
Same-run pairs1
First looks6
Judge anchor checkNot set up yetStarts with the November run
Same model, month to monthNot reported yetWaits for the noise-floor measurement
Generation pairs
Predecessor Successor Too small to count 95% interval on the change
Moonshot AI · Kimi line · lineage proposed, confirm before publishing Human-centered Without guidance Improved
Need revision, 0 to 60% Kimi K3 22 of 68 (32.4%), Kimi K3.1 Pro 7 of 68 (10.3%) Change in points, −40 to +40 −22.1 points, interval −35.3 to −10.3. Counts only beyond ±10. Human-centered Human-centered guidance No detectable change
Need revision, 0 to 60% Kimi K3 2 of 68 (2.9%), Kimi K3.1 Pro 2 of 68 (2.9%) Change in points, −40 to +40 0.0 points, interval −5.9 to +7.4. Counts only beyond ±5. Trauma-informed Without guidance No detectable change
Need revision, 0 to 60% Kimi K3 8 of 24 (33.3%), Kimi K3.1 Pro 5 of 24 (20.8%) Change in points, −40 to +40 −12.5 points, interval −37.5 to +8.3. Counts only beyond ±10. Trauma-informed Trauma-informed guidance No detectable change
Need revision, 0 to 60% Kimi K3 0 of 24 (0.0%), Kimi K3.1 Pro 0 of 24 (0.0%) Change in points, −40 to +40 0.0 points, interval 0.0 to 0.0. Counts only beyond ±5. What this run can see
With 68 drafts per model per condition, a change has to be about 20 points to show reliably when about 30% of drafts need revision,
and about 7 points when about 2% do. So the Benchmark can catch a generation that falls apart under guidance. It can't separate two models
that both sit near zero, and it says so instead of calling a winner.
| Model | Lab | Snapshot | Days before the run | Generation verdict |
| DeepSeek V4.1 Flash | DeepSeek | Oct 17, 2026 | 15 | No tested predecessor; placement only |
| Gemini 4 Argon | Google | Oct 13, 2026 | 19 | No tested predecessor; placement only |
| GPT-6.1 Terra | OpenAI | Oct 26, 2026 | 6 | No tested predecessor; placement only |
| Kimi K3.1 Pro | Moonshot AI | Oct 23, 2026 | 9 | Paired with Kimi K3, same run |
| Muse Wonder | Meta | Oct 12, 2026 | 20 | No tested predecessor; placement only |
| Nemotron 3.5 Super VL | NVIDIA | Oct 15, 2026 | 17 | No tested predecessor; placement only |