SEABED - Dialectal Speech Comprehension (Open-Ended): leaderboard
SEABED (Rajgarhia et al., 2026) is an audio-first reasoning benchmark on Southeast Asian speech built from openly available corpora. In this task the model answers content questions over Thai regional-dialect speech in Pattani Malay, Korat and Khummuang; answers are open-ended and a Grok 4.5 judge marks each answer correct or incorrect, and the score is accuracy in percent, higher is better.
Metric: Accuracy (%; 504 comprehension questions over Thai regional-dialect speech items; open-ended answers judged by Grok 4.5, binary correct or incorrect (graded 0 to 1 for long-form); audio at 16 kHz). Source: arxiv.org. Saturation forecast: Around August 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.5 Flash | 48.7 | #98 |
| 2 | Gemini 2.5 Pro | 44.3 | #262 |
| 3 | Gemini 3.6 Flash | 42.5 | #72 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=seabed-dialectal-speech-comprehension-open-ended · How It Works · Data refreshed daily, snapshot 2026-11-12.