Zoom Scribe
Zoom Scribe

Zoom Scribe: API-Anbieter-Benchmarking & Analyse

Modellentwickler:Zoom
Lizenz:Proprietär
Besuchen
Analyse der API-Anbieter von Zoom Scribe anhand von Leistungskennzahlen, einschließlich Artificial Analysis Wortfehlerratenindex, Geschwindigkeit und Preis.

Highlights

AA-WER v2 · % of words transcribed incorrectly · Lower is better
Input audio seconds transcribed per second · Higher is better
USD per 1000 minutes of audio · Lower is better

Artificial Analysis Wortfehlerratenindex (AA-WER) nach API

Artificial Analysis Wortfehlerratenindex (AA-WER) nach API

% of words transcribed incorrectly · Lower is better · AA-WER v2 incorporates 3 datasets: AA-AgentTalk (50%), VoxPopuli-Cleaned-AA (25%), Earnings22-Cleaned-AA (25%)
Hinweis: Wenn ein Modell bei Earnings22 aufgrund von Zeitlimits Audio in voller Länge nicht zuverlässig verarbeiten kann, teilen wir es in Segmente von ca. 9 Minuten auf (relevant für: Nova 2 Pro, Amazon, GPT-4o Transcribe, OpenAI und GPT-4o Mini Transcribe, OpenAI). Bei Modellen mit noch kürzeren Zeitlimits teilen wir es in Segmente von ca. 30 Sekunden auf (relevant für: Canary Qwen 2.5B, NVIDIA).

Measures transcription accuracy across 3 datasets to evaluate models in real-world speech with diverse accents, domain-specific language, and challenging channel & acoustic conditions.

AA-WER is calculated as an audio-duration-weighted average of WER across ~8 hours from three datasets: AA-AgentTalk (50%), VoxPopuli-Cleaned-AA (25%), and Earnings22-Cleaned-AA (25%). See methodology for more detail.

API-Benchmarks

Artificial Analysis Wortfehlerratenindex vs. Preis

% of words transcribed incorrectly · Lower is better · AA-WER v2 incorporates 3 datasets: AA-AgentTalk (50%), VoxPopuli-Cleaned-AA (25%), Earnings22-Cleaned-AA (25%) · USD per 1000 minutes of audio
Most attractive quadrant

Measures transcription accuracy across 3 datasets to evaluate models in real-world speech with diverse accents, domain-specific language, and challenging channel & acoustic conditions.

AA-WER is calculated as an audio-duration-weighted average of WER across ~8 hours from three datasets: AA-AgentTalk (50%), VoxPopuli-Cleaned-AA (25%), and Earnings22-Cleaned-AA (25%). See methodology for more detail.

Estimated cost in USD to transcribe 1,000 minutes of audio, normalized across providers with different billing models, and including billed reasoning tokens where available. Further detail on the methodology page.

Geschwindigkeitsfaktor

Geschwindigkeitsfaktor

Input audio seconds transcribed per second · Higher is better

Audio file seconds transcribed per second of processing time. Higher factor indicates faster transcription speed. Reported Speed Factor values are medians across benchmark trials from the last 7 days; over-time chart points are daily medians. Artificial Analysis measurements are based on an audio duration of 10 minutes. Speed Factor may vary for other durations, particularly very short durations under 1 minute.

Preis

Preis der Transkription

USD per 1000 minutes of audio

Estimated cost in USD to transcribe 1,000 minutes of audio, normalized across providers with different billing models, and including billed reasoning tokens where available. Further detail on the methodology page.

Zusammenfassung wichtiger Kennzahlen & weitere Informationen

Anbieter
Weitere Einzelheiten
Qwen3.5 Omni Flash
Logo von Qwen3.5 Omni FlashAlibaba Cloud
13.5%
78.5
0.00
Qwen3.5 Omni Plus
Logo von Qwen3.5 Omni PlusAlibaba Cloud
3.5%
95.2
0.00
Nova 2 Pro
Logo von Nova 2 ProAmazon Bedrock
4.9%
23.0
3.10
Amazon Transcribe
Logo von Amazon TranscribeAmazon Bedrock
4.1%
16.9
6.00
Universal-3 Pro
Logo von Universal-3 ProAssemblyAI
3.1%
111.7
3.50
Universal, AssemblyAI
Logo von Universal, AssemblyAIAssemblyAI
3.8%
123.1
2.50
MAI-Transcribe-1.5
Logo von MAI-Transcribe-1.5Microsoft Azure
2.4%
183.3
6.00
MAI-Transcribe-1
Logo von MAI-Transcribe-1Microsoft Azure
2.6%
67.1
6.00
transcribe-03-2026
Logo von transcribe-03-2026Cohere
4.6%
118.7
0.00
Nova-3
Logo von Nova-3Deepgram
5.2%
541.1
4.30
Scribe v2
Logo von Scribe v2ElevenLabs
2.2%
57.0
3.67
Solaria-1, Gladia
Logo von Solaria-1, GladiaGladia
4.1%
80.0
10.17
Solaria-3, Gladia
Logo von Solaria-3, GladiaGladia
3.2%
62.8
10.16
Gemini 3.1 Pro Preview (High)
Logo von Gemini 3.1 Pro Preview (High)Google
2.8%
7.0
18.15
Gemini 3.1 Pro Preview (Low)
Logo von Gemini 3.1 Pro Preview (Low)Google
3.6%
7.4
7.72
Gemini 3 Flash (High)
Logo von Gemini 3 Flash (High)Google
2.9%
16.6
13.70
Gemini 2.5 Flash Lite
Logo von Gemini 2.5 Flash LiteGoogle
5.2%
69.0
6.56
Gemini 2.5 Flash
Logo von Gemini 2.5 FlashGoogle
5.1%
74.5
6.66
Gemini 2.5 Pro
Logo von Gemini 2.5 ProGoogle
2.9%
13.3
11.39
Gemini 3.1 Flash-Lite Preview (Minimal)
Logo von Gemini 3.1 Flash-Lite Preview (Minimal)Google
3.4%
82.3
5.83
Gradium Speech-to-Text
Logo von Gradium Speech-to-TextGradium
6.8%
2.3
13.00
Grok Speech to Text, SpaceXAI
Logo von Grok Speech to Text, SpaceXAISpaceXAI
4.0%
231.1
1.67
Voxtral Mini Transcribe 2
Logo von Voxtral Mini Transcribe 2Mistral
3.6%
85.1
3.00
Voxtral Small
Logo von Voxtral SmallMistral
2.8%
66.2
4.00
Voxtral Mini
Logo von Voxtral MiniDeepInfra
3.8%
78.1
1.00
Modulate STT Batch English VFast
Logo von Modulate STT Batch English VFastModulate
4.2%
64.9
0.42
Parakeet TDT 0.6B V3, Togetherai
Logo von Parakeet TDT 0.6B V3, TogetheraiTogether AI
4.5%
292.8
1.50
Canary Qwen 2.5B, NVIDIA
Logo von Canary Qwen 2.5B, NVIDIAReplicate
4.3%
5.7
0.74
Parakeet TDT 0.6B V2, NVIDIA
Logo von Parakeet TDT 0.6B V2, NVIDIANVIDIA
6.4%
98.9
0.00
Parakeet RNNT 1.1B
Logo von Parakeet RNNT 1.1BReplicate
5.4%
6.3
1.91
GPT Transcribe, OpenAI
Logo von GPT Transcribe, OpenAIOpenAI
3.3%
38.9
4.50
GPT-4o Transcribe
Logo von GPT-4o TranscribeOpenAI
4.0%
38.1
6.00
GPT-4o Mini Transcribe
Logo von GPT-4o Mini TranscribeOpenAI
4.5%
43.5
3.00
Smallest AI Pulse Pro
Logo von Smallest AI Pulse ProSmallest.ai
2.4%
274.6
4.00
Resonant-1
Logo von Resonant-1Reson8
3.4%
306.8
3.60
Rev AI
Logo von Rev AIRev AI
5.9%
13.4
3.33
Smallest AI Pulse
Logo von Smallest AI PulseSmallest.ai
4.4%
278.7
5.00
Soniox v5 Async
Logo von Soniox v5 AsyncSoniox
3.8%
19.0
1.66
Soniox V4
Logo von Soniox V4Soniox
3.9%
19.4
1.66
Speechmatics Melia
Logo von Speechmatics MeliaSpeechmatics
4.9%
210.2
4.00
Speechmatics Standard
Logo von Speechmatics StandardSpeechmatics
5.1%
104.5
7.50
Speechmatics Enhanced
Logo von Speechmatics EnhancedSpeechmatics
4.0%
71.8
12.50
Whisper Large v3 Turbo
Logo von Whisper Large v3 TurboGroq
4.6%
117.1
0.67
Wizper Large v3
Logo von Wizper Large v3fal.ai
4.7%
154.6
0.50
Incredibly Fast Whisper
Logo von Incredibly Fast WhisperReplicate
5.7%
55.4
1.49
Whisper Large v3
Logo von Whisper Large v3Replicate
10.1%
2.7
4.23
Whisper Large v3
Logo von Whisper Large v3fal.ai
4.1%
89.2
1.15
Whisper Large v3
Logo von Whisper Large v3Together AI
4.5%
311.8
1.50
Whisper Large v2
Logo von Whisper Large v2OpenAI
4.1%
29.6
6.00