Nejumi 4 - jaster (0-shot) - JEMHopQA: leaderboard
JEMHopQA, Japanese explainable multi-hop question answering, from the llm-jp-eval jaster suite as run 0-shot by Nejumi Leaderboard 4 (Weights & Biases Japan). Score is character-level F1 x100 on Nejumi's sampled test split; higher is better.
Metric: Character F1 (x100). Source: nejumi.ai. Saturation forecast: Around March 2027. 145 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 72.03 | #64 |
| 2 | Gemini 3.5 Flash | 71.59 | #66 |
| 3 | Gemini 3 Pro (Preview) | 71.06 | #71 |
| 4 | Gemini 3 Flash (Preview) | 69.69 | #90 |
| 5 | Gemini 3.6 Flash | 69.06 | #57 |
| 6 | GPT-5.4 (2026-03-05) (xHigh) | 68.76 | #63 (GPT-5.4 (2026-03-05)) |
| 7 | GPT-5.4 (2026-03-05) (High) | 68.37 | #63 (GPT-5.4 (2026-03-05)) |
| 8 | Gemini 2.5 Pro | 67.94 | #159 |
| 9 | Muse Spark Open (xHigh) | 67.88 | #42 (Muse Spark Open) |
| 10 | Claude Opus 4.5 (Thinking) | 66.01 | #87 (Claude Opus 4.5) |
| 11 | GPT-5.2 (2025-12-11) (xHigh) | 65.97 | |
| 12 | GPT-5 (2025-08-07) (High) | 65.91 | |
| 13 | GPT-5.6 Sol (Max) | 65.84 | #22 (GPT-5.6 Sol) |
| 14 | O3 (2025-04-16) (High) | 65.73 | #125 (O3 (2025-04-16)) |
| 15 | GPT-5.6 Terra (Max) | 64.61 | #40 (GPT-5.6 Terra) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=nejumi-4-jaster-0-shot-jemhopqa · How It Works · Data refreshed daily, snapshot 2026-11-05.