TurboScribe

Three things: the number of files, their length and their place in the queue. The pricing page lists TurboScribe Free as 3 transcripts a day, each file up to 30 minutes long, one upload at a time, with lower priority, so free files wait longer before transcription starts. The homepage adds that no credit card is required, and the getting-started guide walks through a first transcript on a free account.

Some features are listed only on the paid card: translation, bulk exports, all transcription modes and unlimited storage. The page does not say which modes, export formats or how much storage a free account gets, nor when the daily count resets. In practice the 30-minute cap matters more than the daily count for meetings: an hour-long call must be cut into two files before upload.

The homepage FAQ says there are no caps on overall usage and that the only rule is not sharing a login. The high-volume guide sets out what happens at the top end: accounts are sized for at least 720 hours of audio a month, and above roughly 24 hours a day on average an account moves to High Volume Mode. There, files are usually transcribed one at a time, turnaround is no longer the priority, and uploads may be refused with “Try again later” until capacity frees up. The Terms of Service also allow the company to disable an account temporarily if usage significantly exceeds that of other customers.

For more than that, an Unlimited Max article from December 2025 describes a $200 a month plan with dedicated GPU capacity and no upload limits. Organisations can buy one Unlimited licence per user at $120 a year under a single account, according to the Teams post. Neither option appears on the public pricing page.

Yes, if the meeting has already been recorded. TurboScribe works on files you upload through the web interface, and the help center states that there is no API or automated access. No page checked describes joining Zoom, Teams or Google Meet calls, live transcription or automatic meeting summaries. If a notetaker that joins calls is the requirement, that is the job covered on the Otter.ai and Fireflies.ai profiles.

For a recorded call, the relevant settings sit under “Speaker Recognition & More Settings” at upload. You can enter the number of speakers or choose automatic detection, which the help center says can overestimate how many people spoke. Restore Audio is meant for poor recordings and adds 2 to 3 minutes per hour of audio. The homepage says the service is powered by Whisper and offers three modes, with Whale, the most accurate, as the default; it does not name a model version, and its 99.8% accuracy figure comes without a published method. Results can be downloaded as DOCX, PDF, TXT, CSV or SRT and VTT subtitles.

The Security and Privacy FAQ, published March 18, 2025, makes four commitments: files and transcripts are encrypted at rest with algorithms such as AES-256, only the account holder can see them unless shared, transcription runs in-house rather than through third-party transcription APIs, and media files and transcripts are not used to train AI or machine learning models. The primary data infrastructure is in the United States.

Retention depends on what you delete. The Privacy Policy, last updated August 23, 2022, says content is kept while the account is active, and deleted content becomes inaccessible at once and is purged from backups within 90 days. The terms say that after account cancellation content is removed from active systems within 7 days and from backups within 90 days. The policy predates the training statement, and none of the pages checked mentions HIPAA, a data processing agreement or a security certification, so law and health users should ask before uploading client recordings.

The terms and privacy policy name Leif Erikson Ventures, LLC, a U.S. company, but give no postal address. The homepage FAQ introduces the founder as Leif, who says he spent almost a decade building AI systems at companies such as Meta; support runs through support@turboscribe.ai.

On content, the terms say the company claims no intellectual property rights over the material you provide and that uploaded materials remain yours. They do not address commercial use of transcripts separately. The limits are on the service itself: uploads must comply with U.S. copyright law, accounts must belong to a human, access may not be resold, and the service is provided as is, with no service-level agreement for most products.

Cancel from Account Settings, then Manage Subscription and Cancel Plan, or through the payment portal. The homepage FAQ says access continues until the end of the current billing period. The terms add that cancellation takes effect at the end of the paid term unless stated otherwise, that deleting your account typically cancels all subscriptions immediately, and that an email or phone request does not count as cancellation.

Money already paid stays paid. The terms state that all purchases are non-refundable and that unused time in the last billing cycle is not prorated; they mention no exception for local consumer law. That makes the yearly plan a twelve-month commitment from the day of upgrade, so the free plan is the place to check results first.

1. Write a fictional dialogue of about two minutes between two invented people planning a neighbourhood book swap, with no real names, addresses or other personal details, and record it with two voices.
2. Upload it on a free account with the audio language set to English, Speaker Recognition set to 2 speakers, and the default Whale mode.
3. Note how long the file waits before transcription starts and whether the two speakers are labelled correctly.
4. Download the transcript as DOCX and SRT, and note whether translation is offered on the free account.
5. Delete the file and confirm it no longer appears in the dashboard.

Accept if both speakers are separated and both exports open cleanly. Reject if speaker turns are merged or the file stays visible after deletion. No result is recorded here.