Fathom alternative: we could not read Fathom's site, so here is the honest version

What we compared, and how

Everything in the other column was read from public pages on 21 October 2026. Where a page could not be read, the cell says “not measured” — we do not fill those cells from advertising or from memory. Our own column comes from our pricing page, our service contract and our privacy policy, all published on this site.

Where Fathom is better

• We do not know yet, and we would rather say so than pretend. On 21 October 2026 every path of fathom.video — the home page, /pricing, /sitemap.xml and /robots.txt — answered HTTP 451 for us, both directly and through a text proxy.
• If the category is right, a meeting notetaker records the call for you and writes the summary, while our service only ever sees a file you hand it.
• A service that blocks automated readers cannot be compared with us on price, languages or limits; that gap may close, and we will publish what we can read when it does.

Where Transcrib Si is better

• Everything on this page is published by us: the price ($10 a month paid yearly, $120 a year, or $20 a month), the free limits (3 files a day, 30 minutes and 64 MB per file), the language count (Russian and English), the export list (TXT, Markdown, SRT, VTT, CSV, DOCX) and the file limit (120 minutes, 64 MB) on every plan.
• A developer surface: REST v1 under api.transcrib.si, an MCP server with 4 tools, and plugins for Claude and ChatGPT.
• A published privacy policy: which file is processed where, which third party sees a purely English file, how long anything is kept and when it is deleted.
• Openness to machines: llms.txt, llms-full.txt and JSON-LD with our prices and limits, plus a robots.txt that welcomes the AI crawlers.
• No bot in your call. Nothing is recorded unless you press the button.

Who should pick which

• Pick Fathom if you want the meeting recorded and summarised for you — but read the current terms and prices on their site first, because we could not read them on 21 October 2026.
• Pick Transcrib Si if you want a service whose price, limits and data handling are published in full and can be read by a person or an assistant without a login.
• Pick Transcrib Si if the recording already exists and the words are what you need.
• Pick either if you need Russian and English: our engine is tuned for Russian and for mixed Russian-English speech.

Why we are publishing a page with empty cells

Most comparison pages in this niche fill every cell. We do not, because a cell we did not read is not a fact. Our own survey of 21 October 2026 records Fathom as an AI meeting notetaker — that is the market category we found for it — and at the same time records that the site could not be opened: HTTP 451 on every path, including through a text proxy and a headless browser.

That has a practical consequence for you. If you are considering moving from Fathom, the numbers that decide the move — the price of the free plan, the limit on the paid plan, the number of languages, the export formats, what happens to your recording — are all on Fathom's side of the table, and they are all unknown to us as of 21 October 2026. The right move is to read them yourself, or to ask Fathom, and to check the answers against the published facts on our side.

There is a second reason to keep this page. A service that answers 451 to automated readers cannot be read by an assistant either. Our own answer is published so that any person or any agent can read it without a login: llms.txt, llms-full.txt, JSON-LD on every page, and a robots.txt that welcomes the AI crawlers.

What to ask Fathom before you switch

These are the questions we would ask, and we publish them because the answers decide whether moving is worth it:

• What does the free plan include, and what is the limit on the recording length and the file size?
• What does the paid plan cost per month, per seat, and what allowance does it include — minutes, meetings, or hours?
• How many languages does it transcribe, and is there a number for Russian and for mixed Russian-English speech?
• Which formats can a transcript be exported to, and can the export include timecodes and speaker labels?
• Where is the audio processed, which third parties see it, how long is it kept, and what happens to it when the account is closed?
• Is there an API, and are there limits on it?

What we publish on our side

The table at the top of this page is our answer to the same questions, and it is the one we can stand behind: Free at $0 with 3 files a day, up to 30 minutes and 64 MB per file, timestamps and speaker labels, exports to TXT, Markdown, SRT, VTT and CSV, and no card. Pro at $10 a month paid yearly ($120 a year) or $20 a month paid monthly, with a fair use of 1,500 minutes of audio per calendar month, every export format including DOCX, priority queue, full history, REST API keys and MCP access. Team on request. The API is metered by volume, with no subscription attached to it.

On the language question we are narrow and precise: Russian and English, with automatic detection when you do not set the language, and an engine tuned for the mixed Russian-English conversation. On the data question we are explicit: a Russian or mixed file is processed on our own transcription infrastructure outside the United States, a purely English file may be passed to a third-party cloud speech provider in the United States when that fallback is enabled on the server, nothing is sold, nothing is used to train models for other customers, Free transcripts are kept for 7 days, and a deleted file leaves the live system immediately.