VoiceLongMemEval - Indirect 140Q - Descriptive with Hint: leaderboard
Metric: Accuracy (%). Source: arxiv.org. Saturation forecast: Around July 2028. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.8 | 63.1 |
| 2 | Claude Sonnet 4.6 | 59.1 |
| 3 | Qwen 3.5 122B A10B | 57.1 |
| 4 | GPT-5.5 | 56.2 |
| 5 | Qwen3-Next-80B | 51.4 |
| 6 | Gemma 3 12B | 44.5 |
| 7 | Llama 3.3 70B | 42.4 |
| 8 | Llama 4 Maverick | 28.6 |
Interactive version: theaggregate.ai/benchmark?slug=voicelongmemeval-indirect-140q-descriptive-with-hint · How It Works · Data refreshed daily, snapshot 2026-09-24.