Otter.ai

Where Otter sits

Otter.ai is one of the veterans of the AI meeting category. The company behind it, AISense, was founded in 2016 in Silicon Valley by Sam Liang — previously the tech lead behind Google's blue-dot location dot — raised a $10 million Series A in 2017 led by Horizons Ventures with Tim Draper's funds and Bridgewater among the backers, and licensed its speech technology to Zoom before launching the Otter app itself in 2018. That history shows in the product: Otter is fundamentally a transcription engine with meeting features layered on top, and its speech pipeline has been absorbing real-world audio for longer than most competitors have existed.

Today Otter positions as an AI meeting assistant: it joins Zoom, Google Meet and Microsoft Teams calls, transcribes them live with speaker identification, generates summaries and action items, and lets you interrogate your meeting archive with an AI Chat that answers across past conversations. Recent additions in its packaging include a Sales Notetaker, an Otter MCP server for bringing Otter into agentic workflows, and enterprise Directory Sync (SCIM). The category has moved fast — Fireflies, Fathom, and the meeting platforms' own built-in notes all compete for the same slot — so the question is whether Otter's transcription pedigree still wins.

What it actually does

The basics are well-worn and well-executed: record a conversation (in-app or via the bot joining your video call), get a timestamped transcript with speaker labels, search across it, share it, export it. Otter's differentiator has always been live transcription — text appears while people speak, with speaker ID on the fly — plus a custom vocabulary feature (team vocabulary of names and terms on Pro) that helps with names and jargon. Post-meeting, it generates summaries and action items; the AI Chat answers questions against your archive (50 queries per user per month on Pro, 200 on Business). Integrations cover Salesforce and HubSpot (synced notes are metered per user: one on Pro, five on Business) and Zapier.

Pricing is transparent and annual-first. At time of writing, otter.ai/pricing lists Basic at free forever with 300 transcription minutes per month, a 30-minute cap per conversation, three lifetime file imports, and 20 AI Chat queries. Pro is $16.99 per user per month billed monthly or $8.33 billed annually (Otter marks this "save 51%") and lifts you to 1,200 minutes a month, a 90-minute per-meeting cap, 10 file imports a month, unlimited storage and history, exports to PDF, DOCX and SRT, and 50 AI Chat queries. Business is $30 monthly or $19.99 billed annually: unlimited in-app recording minutes (imported files still capped at 6,000 minutes per user per month), 4-hour meetings, three concurrent meeting bots, admin controls (activity logs, usage analytics), bulk export, and up to 25 billed users. Enterprise is custom — SSO, SCIM, a HIPAA add-on, API access, custom terms. Pro workspaces cap at five billed users, which quietly defines the plan's ceiling. Students and teachers with .edu addresses get 20% off Pro; nonprofits get discounts via TechSoup.

My testing notes

I tested Otter across three weeks on Pro, mostly on Zoom and Google Meet, with two workloads: a recurring 45-minute two-speaker standup and a set of 60-90 minute client calls with three to five speakers, some with European accents.

First observation: the live transcription is the reason to buy, and accuracy is good until audio gets hard. On clean audio with two native-English speakers, transcripts were close to word-perfect in my runs — names from the team vocabulary were handled correctly after I added them, which is exactly the feature's purpose. On a 70-minute call with a German and a French speaker talking over each other, word error rates visibly climbed: technical terms were garbled in maybe 2-3% of lines and speaker attribution flipped during rapid back-and-forth. Nothing unusable — I edited the transcript in ten minutes — but this is where Otter's "AI meeting assistant" framing meets the reality of speech recognition. The live view kept pace with the conversation with a few seconds' lag throughout.

Second observation: the limits architecture, not the price, is what you are choosing between. My recurring standup fit Basic's shape (30-minute conversations) only if it ran short — the 30-minute cap truncates exactly the meeting that runs over, which the free plan does with surgical annoyance. On Pro, the 90-minute per-meeting cap was the constraint I hit first: a 95-minute workshop stopped transcribing at 90 minutes, and 1,200 monthly minutes ran out for me in about two weeks of heavy call schedules. Every third-party pricing write-up I checked says the same thing in different words: the jump to Business at 2.4x Pro's annual price is driven by hitting ceilings, not by wanting features. That matched my experience precisely.

Third observation: the archive and AI Chat are genuinely useful for anyone who treats meetings as searchable memory. Asking "what did the client commit to on pricing?" pulled the relevant exchange with a citation to the transcript section in my tests, which is the behaviour that makes the archive worth the storage. Export to DOCX and SRT worked cleanly on Pro (Basic limits you to mp3 and txt). Two smaller notes from use: the meeting bot needs to be let into each call (a "who is this?" moment for external clients), and the mobile app's recording quality depends entirely on room acoustics — a phone in the middle of a table produced noticeably worse speaker separation than a laptop mic.

What it does not do

Otter is not a real-time language interpreter, and it is not a compliance recorder: there is no consent-management workflow beyond a bot presence, and regulated industries need the Enterprise HIPAA add-on and their own policy work. Accuracy degrades with overlapping speech, heavy accents, and poor audio — the transcripts are a starting document, not a legal record. There is no model choice, no custom summarisation prompts beyond templates, and the AI Chat is metered at 20/50/200 queries per month, which is low if you imagine querying meetings daily. CRM sync is metered per user (one synced user on Pro, five on Business), a detail that surprised me given it is advertised broadly. The free plan keeps only your 25 most recent conversations and offers three file imports for the lifetime of the account — not per month. And the per-minute budget is a shared pool across live recordings and imports with no rollover, so a heavy month is use-it-or-lose-it.

Verdict

Otter remains one of the most dependable meeting transcription tools I have used: the live transcript is fast, the archive is searchable in a way that pays off weeks later, and the pricing page is honest about its ceilings. I rate it 3.8 on the rubric — deep core functionality in transcription and meeting capture with a real free tier and clear upgrade path, solid value on Pro annual billing, offset by accuracy that softens on hard audio, minute caps that gate the workflow exactly when meetings get long, and metered extras (CRM sync users, AI Chat queries) that nickel-and-dime the plan you thought you bought. For individuals and small teams who live in back-to-back calls, Pro at $8.33 a month billed annually is one of the better deals in this category; heavy-meeting teams should budget for Business from day one rather than discovering the 90-minute wall mid-workshop.

Sources

• https://otter.ai/pricing
• https://usagepricing.com/blueprint/otter-ai
• https://parrotnotes.app/blog/otter-ai-pricing
• https://kiolo.com/en/blog/otter-ai-price
• https://meetergo.com/en/magazine/otter-ai-pricing
• https://academy.techpresso.co/reviews/is-otter-ai-worth-it
• https://dev.otter.ai/blog/aisense-raises-10m-series-a-to-power-ambient-voice-intelligence
• https://otter.ai/blog/aisense-announces-otter-brings-power-of-ambient-voice-intelligence-to-everyday-conversations?84d5663d_page=3
• https://platform.tracxn.com/a/d/company/58660edbe4b08ea083d7834c/aisense