Loading…

Voice & speech

Transcribe audio

Refine candidates (optional)

Advanced requirements

Official API requiredMust run locallyMust work entirely offline

Candidates to compare

Based on documented capabilities, not measured performance on your task. Leaderboard context is shown separately.

What to test first

Use the same script or recording. Check names, pronunciation or transcription errors, speaker consistency, and the rights needed for your use.

ElevenLabs · Eleven v3 / Scribe v2

ElevenLabs · elevenlabs.io

Speech transcription[1]Aging
Supported
Scribe v2 timestamps/speakers; separate scribe_v2_realtime model and rate.
Web interface[2]Aging
Supported
Publicly documented channel; account, region, entitlement and quota were not tested.
Official API[3]Aging
Supported
Publicly documented channel; account, region, entitlement and quota were not tested.

Full evidence and sources

No exact leaderboard match

Why this fits

Speech transcription
Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High
Scribe v2 timestamps/speakers; separate scribe_v2_realtime model and rate.
The documented input/output workflow matches this task. Test prompt adherence and output quality with the same brief used for the other candidates.
Web interface
Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High
Publicly documented channel; account, region, entitlement and quota were not tested.
A documented web interface gives you a way to try the workflow in a browser. Check which features and exports your selected plan includes.
Official API
Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High
Publicly documented channel; account, region, entitlement and quota were not tested.
A documented API makes this a candidate for repeatable, programmatic workflows. Check account access, limits, and integration work before committing.

Strengths

Speech transcription:
Supported · Official fact · Aging · View source
Official API:
Supported · Official fact · Aging · View source
Cost:
API v3 USD 0.08 / 1,000 characters; Scribe v2 USD 0.22 / audio hour · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High
Realtime USD 0.39/hour; entity/keyterm add-ons separate. Live page supersedes search snippet USD 0.10. Web credits are not API dollar balance.

Trade-offs

Unreviewed dimensions remain open questions for your trial.

Cost: API v3 USD 0.08 / 1,000 characters; Scribe v2 USD 0.22 / audio hour · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High
Realtime USD 0.39/hour; entity/keyterm add-ons separate. Live page supersedes search snippet USD 0.10. Web credits are not API dollar balance.
Plan-dependent. Check the official price and billing period.
Access: Local deployment:
Unknown
Official API:
Supported · Official fact · Aging · View source
Web interface:
Supported · Official fact · Aging · View source
Workflow: Use the official API, then import the output into Proof Lab.
Commercial use and rights: Unknown
Rights depend on the product, plan, inputs and intended use. Review the linked terms.

Unknowns

Correctness · Consistency

Evidence

Text-to-speech

Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High

eleven_v3, up to 5,000 input characters; previous generation, not latest.

Speech transcription

Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High

Scribe v2 timestamps/speakers; separate scribe_v2_realtime model and rate.

Local deployment

UnknownLast verified: Not reviewedConfidence: Low

Official API

Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High

Publicly documented channel; account, region, entitlement and quota were not tested.

Web interface

Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High

Publicly documented channel; account, region, entitlement and quota were not tested.

Cost

API v3 USD 0.08 / 1,000 characters; Scribe v2 USD 0.22 / audio hour · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High

Realtime USD 0.39/hour; entity/keyterm add-ons separate. Live page supersedes search snippet USD 0.10. Web credits are not API dollar balance.

Commercial use

UnknownLast verified: Not reviewedConfidence: Low

Google Cloud · Chirp 3 STT V2

Google · docs.cloud.google.com

Speech transcription[1]Aging
Supported
V2 chirp_3. Streaming, sync and batch; documentation conflicts on word timestamps and synchronous diarization, so those combinations stay unknown.
Web interface[2]Aging
Supported
Official API[3]Aging
Supported
Publicly documented channel; account, region, entitlement and quota were not tested.

Full evidence and sources

No exact leaderboard match

Why this fits

Speech transcription
Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High
V2 chirp_3. Streaming, sync and batch; documentation conflicts on word timestamps and synchronous diarization, so those combinations stay unknown.
The documented input/output workflow matches this task. Test prompt adherence and output quality with the same brief used for the other candidates.
Web interface
Supported · Official fact · Aging · View source Last verified: 2026-09-19Confidence: High
A documented web interface gives you a way to try the workflow in a browser. Check which features and exports your selected plan includes.
Official API
Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High
Publicly documented channel; account, region, entitlement and quota were not tested.
A documented API makes this a candidate for repeatable, programmatic workflows. Check account access, limits, and integration work before committing.

Strengths

Speech transcription:
Supported · Official fact · Aging · View source
Official API:
Supported · Official fact · Aging · View source

Trade-offs

Unreviewed dimensions remain open questions for your trial.

Cost: Unknown · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: Low
V2 Standard reference USD 0.016/min first 500k; dynamic batch USD 0.003/min. Generic chirp footnote does not independently establish exact chirp_3 mapping; no automatic cost estimate or V1 free-tier transfer.
Plan-dependent. Check the official price and billing period.
Access: Web interface:
Supported · Official fact · Aging · View source
Local deployment:
Unknown
Official API:
Supported · Official fact · Aging · View source
Workflow: Use the official API, then import the output into Proof Lab.
Commercial use and rights: Unknown
Rights depend on the product, plan, inputs and intended use. Review the linked terms.

Unknowns

Correctness · Consistency · Cost

Evidence

Speech transcription

Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High

V2 chirp_3. Streaming, sync and batch; documentation conflicts on word timestamps and synchronous diarization, so those combinations stay unknown.

Web interface

Supported · Official fact · Aging · View source Last verified: 2026-09-19Confidence: High

Local deployment

UnknownLast verified: Not reviewedConfidence: Low

Official API

Supported · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: High

Publicly documented channel; account, region, entitlement and quota were not tested.

Cost

Unknown · Official fact · Aging · View source Last verified: 2026-09-30T17:37:41.000ZConfidence: Low

V2 Standard reference USD 0.016/min first 500k; dynamic batch USD 0.003/min. Generic chirp footnote does not independently establish exact chirp_3 mapping; no automatic cost estimate or V1 free-tier transfer.

Commercial use

UnknownLast verified: Not reviewedConfidence: Low