AppTek Call-Center Dialogues: leaderboard

AppTek Call-Center Dialogues (2026) is a long-form English speech recognition test set of spontaneous role-played call-center conversations in 14 accents, recorded for evaluation only. This board is the word error rate in percent, averaged over the accents, with audio segmented by the Silero voice-activity detector; lower is better.

Metric: Word error rate (%, mean over the 14 accents) on AppTek Call-Center Dialogues (128.6 hours of spontaneous role-played English call-center conversations, 14 accents), default inference settings, scored per session after Open ASR Leaderboard text normalization, audio segmented with the Silero VAD; lower is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 12 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen3-ASR-1.7B8.3
2parakeet-tdt-0.6B-v39.2
3Qwen3-ASR-0.6B9.2
4canary-qwen-2.5B9.2
5parakeet-tdt-0.6B-v29.6
6Phi-4 Multimodal Instruct10#4154
7canary-1B-v211.2
8kyutai-stt-pytorch-stt-2.6B-en11.3
9granite-speech-3.3-8B11.9
10granite-speech-3.3-2B13.3
11Whisper Large-v315
12whisper-large-v216

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=apptek-call-center-dialogues · How It Works · Data refreshed daily, snapshot 2026-11-21.