Free speech-to-text allowances: Otter, Notta, IBM, Rev, Oct 2026
Otter Basic 300 minutes, Notta Free 120, IBM Lite 500, Rev 5 hours of credit. The same minutes on Sume STT cost $1.20 to $5.00. Limits per file.
Free speech-to-text allowances in October 2026 range from 120 minutes a month on Notta to 500 on IBM Lite, and I found none in the Sume docs: the same minutes cost $1.20 to $5.00 at $0.01 a minute. The allowances differ more in their limits than their size, and the limits are what stop most people.
Figures are from the pricing pages of Otter, Notta, IBM and Rev AI, read 2026-10-03.
What each free plan gives
The Rev line is a credit, not a monthly allowance: free credits equal to 5 hours of Reverb ASR for new users.
| Service | Free allowance | Limit that bites | Same minutes on Sume |
|---|---|---|---|
| Otter Basic | 300 minutes a month | 30 minutes per conversation, 3 lifetime imports | $3.00 |
| Notta Free | 120 minutes a month | 3 minutes per recording | $1.20 |
| IBM Lite | 500 minutes a month | No per-file limit stated | $5.00 |
| Rev AI new-user credit | 5 hours of Reverb (300 minutes) | One-time credit | $3.00 |
Which limit matters
Notta Free allows only 3 minutes per recording, so a one-hour call does not fit at all; Pro raises that to 5 hours. Otter Basic caps a conversation at 30 minutes and lifetime file imports at 3, so uploaded recordings run out quickly. IBM Lite's page names no per-file limit, only 500 minutes and 38 models.
What Sume costs instead
I found no free allowance for Sume STT in the docs. Its cap is 600 seconds per job, so 120 minutes is 12 jobs, 300 minutes is 30 jobs and 500 minutes is 50 jobs, each $0.10. For a few hours a month the cash cost is under the price of a coffee, and the benefit is one API for transcripts, clips and captions.
- Casual use under about 300 minutes a month: a free plan is fine.
- Uploaded recordings over 30 minutes: check Notta Free and Otter Basic limits first.
- Building a pipeline: pay the cent a minute and keep one integration. See splitting long recordings.
How many 45-minute meetings fit
A 45-minute meeting is a useful test, because the per-file limits decide whether it fits at all.
| Service | Fits one 45-minute file? | Meetings from the allowance |
|---|---|---|
| Otter Basic | No, 30-minute limit | 0 as one file |
| Notta Free | No, 3-minute limit | 0 as one file |
| IBM Lite | No per-file limit stated | 11 (495 minutes) |
| Rev AI credit | Not stated | 6 (270 of 300 minutes) |
What Sume does and does not do
Sume text to speech is $0.0475 per 1,000 characters, up to 20,000 characters per request. Speech to text is $0.01 per minute of audio, up to 600 seconds per job.
It returns a transcript with timings; it has no meeting bot, no app and no transcript editor.
Sources
Related posts
More in Comparisons
- Full-duplex video AI: what Griffin changes, what Sume does
Full-duplex video AI listens, watches and answers at once. Tavus Griffin is gated; Sume makes scripted avatar clips by job. Where each fits.
- GEMA v Suno ruling: what to check before AI music goes in an ad
A Munich court ruled against Suno on 31 July 2026, not final. What it says, what it leaves open, and a record to keep for any AI track you put in an ad.
- Gemini 2.5 Flash Image sunset Oct 2: what 1,000 replacements cost
Google marks Gemini 2.5 Flash Image ($0.039) deprecated, sunset October 2, 2026. Nano Banana 2 is $0.067 at 1K, so 1,000 images cost $67.
- Gemini app Omni video needs a Google AI plan; the API is separate
Who can make Omni video in the Gemini app, what uploads it accepts, and how that differs from calling gemini-omni-flash-1.1 through Sume's Video Router.
Written by Sume