Fact · Speech recognition
0.019seconds (time to first partial after end of speech)
In Artificial Analysis's June 2026 streaming benchmark, Deepgram Flux had the fastest first partial transcript after end of speech, at 0.019 seconds.
Checked against its source on October 17, 2026
In the source's own words
“Flux leads both Final and Partial at 0.020s and 0.019s respectively”
Announcing AA-WER Streaming (Artificial Analysis), Artificial Analysis, Artificial Analysis, 2026, Key findings.
Open the original- What it measures
- The benchmark used about 8 hours of audio, and its latency starts when a voice activity detector reports the end of speech.
- What it does not show
- One benchmark on one audio set, with a 7.36 percent word error rate for Flux. It does not measure time from the start of speech.