Get This Tool
voxai
Summary
Most meeting transcription tools give you a text dump after the call ends — useful for notes, useless for the negotiation that just went sideways while you were talking. VoxAI feeds the live transcript to your AI as the conversation unfolds, so the advice arrives before the moment passes.
VoxAI runs entirely on macOS, records locally, and splits your mic from system audio on calls so speaker labels come from channel identity rather than voice-guessing. It connects to Claude, Cursor, or any MCP-compatible client via a single copied prompt — no extension installs, no Python environment, no API-key configuration step. The AI reads the growing transcript in a side panel and can surface risks, questions, or draft action items mid-conversation. The wall appears at ten sessions: after the trial, new recordings cap at ten minutes until you unlock the paid tier. Teams who run long client calls or multi-hour workshops hit that ceiling fast.
Bottom line: Pick this for macOS professionals who need live AI advice during calls and want recordings that never touch a server — but if your team is on Windows or runs sessions longer than ten minutes before they can evaluate a purchase, the platform and trial structure will stop you before you start.
Pricing Plans
Free- Price
- $19.99 one-time or $19.90 direct
- Free Tier
- Ten sessions; then new recordings capped at 10 minutes and live copilot paused
Trial
Ten full sessions of the complete app
- Full transcription and speaker labels
- Live AI copilot
- All features unlocked
Unlock (App Store)
One-time in-app purchase
- Unlimited recordings
- Permanent unlock
- Restore on same Apple Account
Direct (.dmg)
One-time purchase via direct download
- Permanent local license
- Two Macs
- Free updates through 1.x
View full pricing on voxai.ethanflow.com →
Pricing may have changed since last verified. Check the official site for current plans.
Community Performance Report Card
No community ratings yet. Be the first to rate this tool!
Community Benchmarks Community
Sign in to submit a benchmarkNo community benchmarks yet. Be the first to share a real-world data point.
Pros
Sign in to edit- Live transcript streaming to your AI client during the conversation, so advice on a negotiation risk or a contract clause arrives while you can still act on it — not in the debrief.
- Channel-based speaker identification on calls (mic vs. system audio), which means labels are accurate even when voices are similar, so you avoid the manual cleanup that ruins a transcript's usefulness.
- Fully local recording and transcription with cloud explicitly opt-in, so sensitive client conversations, legal interviews, or internal strategy calls never leave the machine by default.
- One-time purchase with no account required, so there is no subscription to manage and no credential to rotate when someone leaves the team.
- MCP connection via a single copied prompt — no extension, no environment setup — which means an AI client like Claude or Cursor is live in under two minutes, and the setup does not break on macOS updates.
Cons
Sign in to edit- The trial caps at ten sessions before new recordings are limited to ten minutes — teams running long client calls, extended interviews, or multi-hour workshops will hit the ten-minute ceiling before they can fully evaluate whether the tool fits their workload, which pushes the purchase decision earlier than the trial design implies.
- macOS 14 and Apple Silicon are required with no Windows or Linux support described anywhere in the vendor materials — cross-platform teams cannot standardize on this tool, and members on non-Apple hardware must use a different transcription workflow entirely, which usually means the team abandons VoxAI for a browser-based alternative like Otter or Fireflies that runs on any OS.
- There is no API exposed and the tool is not agentic — it streams context to your AI but does not act on it autonomously. Teams who want the AI to automatically file action items, send follow-up emails, or update a CRM from the transcript must build that layer themselves in their AI client, which means the automation work lands outside VoxAI.
Community Reviews
Sign in to write a reviewNo reviews yet. Be the first to share your experience.
About
- Platforms
- macOS (Apple Silicon, 14+)
- API Available
- No
- Self-Hosted
- Yes
- Last Updated
- 2026-07-27T08:16:26.419Z
Best For
Who it's for
- macOS users in meetings or calls needing transcription and AI support
- Professionals wanting local data control and one-time purchase
- Teams using Claude, Cursor, or similar MCP-compatible clients
What it does well
- Live meeting transcription with speaker labels and AI notes
- Call recording and analysis with channel-based identification
- In-person interviews or negotiations with real-time AI advice
- Voice interaction with AI for spoken queries and responses
Integrations
Discussion Community
Sign in to commentNo discussion yet. Sign in to start the conversation.
Spotted incorrect or missing data? Join our community of contributors.
Sign Up to ContributeCommunity Notes & Tips Community
Sign in to contributeBe the first to contribute. General notes, observations, gotchas, and tips from people who use this tool day-to-day.
Frequently Asked Questions
- Is voxai free?
- voxai has a permanent free tier alongside paid upgrades (paid plans from $19.99 one-time or $19.90 direct). You can keep using a baseline version indefinitely without paying.
- Is voxai open source?
- Yes. voxai is open source.
- Can I self-host voxai?
- Yes. voxai supports self-hosting on your own infrastructure.
- What platforms does voxai support?
- voxai is available on: macOS (Apple Silicon, 14+).
Hours Saved & ROI Stories Community
Sign in to contributeBe the first to contribute. Concrete time/cost savings, with context. e.g. "Cut my code review backlog from 4h to 45m per week."
Curated lists that include this category
Most transcription tools operate on a record-then-review loop: you finish the meeting, upload the file, wait for the transcript, then ask your AI what happened. VoxAI breaks that loop by maintaining a live transcript stream — speaker-labeled, locally processed — and handing it to whatever AI client you already use via MCP over stdio. The workflow is three steps: drag the app to Applications, paste one connection prompt into your AI client, pick a scene (meeting, call, in-person, or voice chat), and the AI begins reading the conversation as it happens.
The differentiating feature is how speaker identity works. In call mode, VoxAI separates your microphone channel from system audio, so the labels are derived from the source channel rather than acoustic voice-matching — a distinction that matters when two people sound similar or background noise is high. In meeting mode, live speaker labels appear in real time. The vendor states recording, transcription, speaker labeling, and the local voice model can all run on-device with no account required; cloud paths exist but are explicitly opt-in behind a clear switch.
VoxAI fits macOS users in two specific situations: professionals who want a local data guarantee (no account, no default cloud upload, archive stays on-device forever regardless of purchase status) and practitioners already using MCP-compatible clients who want the AI they trust to be present during conversations rather than consulted afterward. It does not fit Windows or Linux users — the vendor states macOS 14 and Apple Silicon are required. It also does not fit teams who need the AI to autonomously act on what it hears; VoxAI streams context to your AI but does not itself plan or execute tasks.
The MCP connection uses stdio with no open network port, which the vendor positions as a security property. The trial delivers the complete app for ten sessions — the vendor states only real recordings of 60 seconds or more count against that limit — after which new recordings cap at ten minutes until a one-time unlock is purchased through the App Store or via direct download.