Fact · Speech recognition
0.020seconds
In Artificial Analysis's June 2026 streaming benchmark, Deepgram Flux was the fastest service on both the first partial and the first final transcript after end of speech, at 0.020 seconds for the final transcript, with a 7.36 percent word error rate.
Checked against its source on October 17, 2026
In the source's own words
“@DeepgramAI Flux leads both Final and Partial at 0.020s and 0.019s respectively (both 7.36% WER)”
Announcing AA-WER Streaming (Artificial Analysis), Artificial Analysis, Artificial Analysis, 2026, Key results, 'Fastest transcription'.
Open the original- What it measures
- Artificial Analysis tested streaming speech-to-text models on about 8 hours of audio, measuring latency from the end of speech as detected by the Silero voice activity detector.
- What it does not show
- Fastest is not most accurate: the same benchmark gives Deepgram Flux a 7.36 percent WER, higher than several rivals. The test is one independent benchmark on its own audio sets, and vendors' prices and models change.