MESH - Egocentric Static Transcript Direction: leaderboard

MESH tests spatial hearing across synchronized receivers and moving speakers. Direction, distance, position and transcription tasks remain separate. Word error rate can exceed 100%. Overall gives equal weight to localization and per-question transcription quality; it is not an average of the printed task metrics. This board reports native localization success (%). Higher is better.

Metric: Native localization success (%). Source: arxiv.org. Saturation forecast: Around May 2027. 10 models tracked. Updated 14 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Qwen3.8 Omni Flash50.03#115
2MESH-agentEvaluation run: V2; Qwen3-14B planner and source specialist tools47.17
3Qwen3 Omni 30B A3B Instruct (Harness: MESH FOA-W)10.15
4Qwen 3.5 Omni Plus (Harness: MESH FOA-W)9.37
5Qwen 3.5 Omni Plus (Harness: MESH FOA-sep)9.15
6Qwen3 Omni 30B A3B Instruct (Harness: MESH FOA-sep)8.71
7MiDasheng-Spatial7.95
8JAEGER Neural7.54
9JAEGER Classical0.19
10BAT0

No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.

Interactive version: theaggregate.ai/benchmark?slug=mesh-egocentric-static-transcript-direction · How It Works · Data refreshed daily, snapshot 2026-11-16.