Sauti / elevenlabs
ElevenLabs Scribe v1
Deprecated by the vendor
WER
13.7%
95% CI 12.4%–15.2%
Rank
#11
of 12 measured
Latency
3.2s
mean, per clip
Cost
—
per audio hour
10 other models are a statistical tie with this one — their confidence intervals overlap, so the ordering between them is not something this data can settle: nova-3, gpt-transcribe, Whisper large-v3 (self-hosted), universal-2, universal-3-5-pro, Whisper small (self-hosted), gpt-4o-mini-transcribe, whisper-1, scribe_v2, gpt-4o-transcribe.
What it is
The first Scribe generation, explicitly deprecated in ElevenLabs' own model docs and carrying no published rate of its own. We publish no price for it rather than borrow v2's.
Pick it when
Nothing recommends it for new work. It is on the board only to measure the v1-to-v2 delta.
Think twice when
Always, for new integrations. Note the delta it exists to show is essentially nil on our corpus: v1 and v2 finished within a rounding error of each other, with near-identical intervals.
Assessment written against the board of 2026-08-08. The figures above are read live from the current published snapshot (2026-08-08), so if those dates differ, trust the figures.
What the vendor publishes
elevenlabs publishes no stated word error rate figure for this model on any public benchmark. That absence is itself the finding. Source (read 2026-08-15).Now listed as DEPRECATED — "first generation speech recognition (outclassed by v2 models)", with scribe_v2 recommended instead. No figure was ever published for it.
How this was measured
250 clips (2.9 hours) of unscripted conversational English from People’s Speech (MLCommons, CC-BY), human-transcribed. Scored on a held-out split (N=77) never used to tune anything, through one fixed text normalizer, with bootstrap 95% confidence intervals. One corpus, one language, one modality: batch pre-recorded. It does not tell you how this model handles your accents, your domain vocabulary, or streaming.
Cost is elevenlabs’s own published pre-recorded pay-as-you-go list price, read from their pricing page on 2026-08-08. Streaming and committed-volume rates differ.Explicitly deprecated by ElevenLabs and carrying no published rate of its own. Kept on the board for the v1-to-v2 accuracy delta; we publish no price rather than borrow the v2 rate.