180 free Speech2Text minutes after registration
Recognition is limited to 15 minutes per day.
Conditions
- Usage beyond the free allowance has a separate per-minute price.

Speech2Text is an online service that converts audio and video into editable text with paragraphs, timestamps, speaker separation, subtitles and summaries.
Offers are ranked by benefit: fixed-amount savings from highest to lowest, then percentage discounts, followed by deals without a numeric saving.
Recognition is limited to 15 minutes per day.
No recently ended offers are shown for this store yet.
Step by step
A concise sequence from choosing an offer to checking its conditions.
Open the offer and register with Speech2Text.
Confirm the free minutes and daily allowance appear in the account.
Prepare a short clean sample without confidential data for the first test.
Upload a file or paste a supported link and select the recognition language.
Enable required features such as speaker separation, timestamps, subtitles or summary.
Run processing and verify names, numbers, terminology and speaker boundaries against the source audio.
Export the result in the required format and delete the source and transcript from the account when no longer needed.
If minutes or a charge are calculated unexpectedly, stop new jobs, save the plan screen and activity history, and contact support before further use. Do not treat unused minutes as a cash balance unless the plan terms explicitly say so.
The current pricing page states there are no automatic charges. Before activating a plan, still verify billing period, included minutes, daily allowance, overage price and business limits, then keep the plan confirmation.
After registration, the free plan grants 180 minutes but limits recognition to 15 minutes per day. Usage beyond the free allowance has a separate per-minute price; check balance and calculation before processing a long file.
The user is responsible for uploaded files, links and transcripts and must not submit prohibited content or infringe others' rights. For speech containing personal, medical, financial or commercial data, assess participant consent and internal retention rules first.
The official page states that files and transcripts are not retained after the user deletes them and that transfer is encrypted. This does not replace user action: after exporting the result, remove unnecessary material from the account.
For recognition issues, include format, duration, language, speaker count, upload method and a timestamped error example. Do not send a full confidential recording when a short anonymised excerpt is enough.
Accuracy depends on the source. Clear speech, low background noise, limited overlap and the correct language normally matter more than recording length. For important publication, always verify names, numbers, terminology and quotations against the audio.
A file can be uploaded or a supported video can be supplied by link. Meetings can use a bot or Chrome extension; the extension can record a tab or screen, microphone and system audio and lets the user choose whether to send the recording to the service.
The transcript can be edited in the browser and exported as DOCX, TXT or PDF, with subtitles in SRT. API, webhooks, queues and batch processing are available for automation, with separate integration terms.
The service suits interviews, lectures, meetings, calls, webinars and subtitle preparation when the user can review the result and has the right to upload the recording and process participants' speech.
Use the free allowance on representative recordings rather than a demo file: measure manual correction, speaker separation quality and actual minutes. Then compare per-minute, monthly and business plans against real volume and user count.
Offers last changed: July 13, 2026
They are granted after registration on the free plan, with recognition limited to 15 minutes per day. Additional use has a separate per-minute price, so check the balance before a long file.
Yes. AI can mishear names, numbers, terminology, accents and overlapping speech. For publication or decisions, verify important passages against the audio even when the overall text looks confident.
The service accepts common audio and video formats and links from supported platforms. Transcripts export as DOCX, TXT and PDF, subtitles as SRT, and the API provides additional machine-readable formats.
Diarisation suffers with overlapping speech, similar voices, echo and weak recording. Check microphones and channels before the meeting, then rename and merge turns manually after processing.
The service states that transfers are encrypted and content is no longer stored after user deletion. Still verify participant consent, contractual limits and internal policy, use an anonymised test and delete unneeded files.
Test representative recordings on the free plan and measure actual minutes and corrections. For one user compare volume, while teams should also compare users, concurrent streams, API and data requirements.