Back to home

Terms of Service

Last updated 2026-09-29 · Available in English only for now

These Terms govern your use of the FlowMic hosted service at flowmic.app (the "Service"), operated by FlowMic (flowmic.app) ("we", "us").

They do not govern the FlowMic software itself. FlowMic is open-source software licensed under the AGPL-3.0. If you download it, build it, or run it on your own hardware, your rights come from that license and these Terms do not apply to you. These Terms apply only when you use *our* servers.

By creating an account you agree to these Terms.

1. What the Service does

FlowMic lets you speak into a phone and have the transcribed text delivered to an input field on your computer. Our hosted service provides:

  • a relay that connects your phone to your computer when they are not on the same network;
  • managed speech recognition and language-model processing — audio and text you choose to process are handled by the engines described in the Privacy Policy (some run by us, some by providers processing on our behalf; we retain neither the audio nor the text), metered as described in §3;
  • an account, device pairing, and the engine configuration you set for your account in the console. Your general preferences (scenario card, custom terms, "AI polish", the language you speak) stay on your phone; the Service does not synchronise them;
  • website voice input — a script you can add to your own website so that its visitors can speak into text boxes on your pages (see §9).

The Service is not required to use FlowMic. On a single network, the software works with no account and no connection to us at all.

2. Your account

You must be at least 16 years old, or the minimum age of digital consent in your country if that is higher.

You are responsible for keeping your password safe and for activity under your account. Tell us promptly if you believe it has been compromised.

One account is for one person. You may pair as many of your own devices as you like. Website voice input (§9) lets visitors to your website use the Service through your account without accounts of their own. That is allowed, and it is the one way other people may use your account.

3. Plans and usage limits

PlanPriceSpeech recognitionLanguage-model processingComputers
Free$020 minutes / month1,000,000 tokens / month2
Pro$12 / month1,000 minutes / month10,000,000 tokens / month3
Max$28 / month3,000 minutes / month50,000,000 tokens / month10

Prices are in US dollars. Any tax due is added at checkout and handled by our payment provider, which sells the subscription to you as merchant of record. A paid plan renews each month until you cancel it, and you can cancel at any time from the console.

The table above shows the price for new subscriptions. As long as your paid plan stays active, it renews at the price you signed up with, even if our published rates change; switching plans or resubscribing after a plan ends uses the price published at that time.

If we add a different billing period later, its price will be published on the pricing page and named here before it can be bought.

The free plan is not a trial. It does not expire, it needs no payment method, and nothing on it stops working when paid plans exist.

Refunds, cancellation, and the 14-day right to withdraw are set out in the Refund Policy, which forms part of these Terms.

Limits reset at the start of each billing cycle (UTC) — based on your registration date (if you have never subscribed), your subscription renewal date (for Pro and Max), or the date your subscription ended (if you are no longer subscribed). Usage is measured as audio duration and as language-model tokens; we do not retain the content that produced those numbers (see the Privacy Policy). Engines you configure with your own credentials, and engines running on your own hardware, are not counted against these limits and are never billed — but when such a session goes through our cloud relay, it can still appear in your usage detail, marked as using your own key. Sessions that stay on your local network are not recorded there at all.

When you reach a limit, cloud transcription stops until your next cycle begins. Nothing you have already created is deleted or locked, and FlowMic keeps working on your own network.

Your free plan does not change because paid plans exist. Its limits are the ones in the table above, and we do not lower them: if we ever change that row it will be upward. You will not be asked for a payment method to keep using it.

4. Acceptable use

Do not use the Service to:

  • break the law, or infringe anyone's rights;
  • record people who have not consented, where consent is required;
  • resell, sublicense, or provide the Service to third parties as your own (offering website voice input to your own website's visitors under §9 is allowed);
  • attack, overload, or attempt to gain unauthorised access to our systems, or circumvent usage limits.

Security research is welcome — see the project's SECURITY.md for how to report findings. Testing conducted in good faith under that policy is not a breach of this section.

We may place an account under restricted use when it is causing harm to the Service or to others. Restricted use means you can still sign in, but product surfaces show a restriction notice and ordinary product actions are refused. You keep the ability to export your data and to close your account from the console. We do not provide an appeal channel for the restriction.

The notice in the product is how you learn that the account is restricted, and it names a reason — one of a short published set: a breach of these Terms, automated or bulk activity, an account-security concern, a legal or regulatory requirement, or a decision following a review. That is the whole of what we tell you: the internal notes our operators write are kept for our own records and are not shown to you, and where the law prevents us from being more specific we will say no more than the reason above.

5. Third-party speech engines

You may configure FlowMic to use a third-party provider (for example OpenAI or Deepgram) with your own credentials. When you do, your audio goes to that provider under their terms and privacy policy, not ours. When your phone reaches your computer through our relay, that audio passes through our server on its way to the provider — we do not inspect or retain it. On your own network, we are not in the path at all. We are not responsible for what a provider you chose does with your data.

If you set up your own engine and also use website voice input (§9), your visitors' audio goes to that engine too, with your credentials. You are then the one who chose that provider for them.

The built-in cloud engines are a different arrangement: there we choose the provider, and it works on our behalf and on our account rather than yours. Who those providers are and what they receive is named in the Privacy Policy under "Who else sees your data" — this section deliberately does not repeat the names, so there is only one place to keep correct when they change.

6. Your content

Whatever you say, transcribe, or store stays yours. We claim no rights to it. The same goes for what visitors say through website voice input: we claim no rights to it, and we keep none of it.

We do not store your transcripts on our servers. What we do store, and for how long, is set out in the Privacy Policy.

7. Availability, changes, and ending the Service

We aim for the Service to be available and reliable, but we do not guarantee uptime. This is a small operation and you should assume outages will happen.

We may change these Terms. If a change materially affects you, we will give at least 30 days' notice by email and in the console before it takes effect. Continuing to use the Service after that means you accept the change.

You may close your account at any time from the console. If we discontinue the Service, we will give at least 60 days' notice and provide a way to export your data before it stops.

8. Warranty and liability

The Service is provided "as is", without warranties of any kind, to the fullest extent the law allows.

To the fullest extent permitted by law, our total liability to you for any claim relating to the Service is limited to the amount you paid us in the 12 months before the claim arose.

Nothing here limits liability that cannot legally be limited — including liability for death or personal injury caused by negligence, or for fraud.

If you are a consumer, you keep all rights that your local consumer law gives you, and nothing in these Terms takes them away.

9. Website voice input

Website voice input lets you put a FlowMic microphone on your own website. Visitors speak into their computer's microphone, or into their phone after scanning a code, and their words appear in text boxes on your page. They need no FlowMic account. You set it up in the console with a site key.

Tell your visitors. Your website's privacy notice must say that voice input on your pages is provided by FlowMic and that visitors' speech is sent to FlowMic to be transcribed. It must also link to our Privacy Policy at flowmic.app/privacy and name the section that covers this ("Speaking into FlowMic on another website"). Our voice input shows its own link to that policy; that link does not replace your notice.

Your key works only on the addresses you list. Each site key must list at least one exact website address (for example https://shop.example), and at most 16. Patterns and wildcards are not accepted; localhost and 127.0.0.1 are accepted for development. List only websites you operate or are authorised to add voice input to. The key is visible in your page's source, so anyone can copy it. The address check stops other websites from using it in ordinary browsers, but it cannot stop every tool. We do not verify who owns a domain. Set a limit on each key if you want to cap what it can spend.

Where not to use it. Do not use website voice input:

  • to collect health, biometric, or other special-category information, information about children, or payment card data, unless you have a legal basis and give any notice the law requires;
  • on websites or pages directed at children under 16;
  • to record anyone without the notice or consent the law requires.

Voice input never writes into password fields.

What it costs you. All speaking by your visitors is counted against your account's plan (§3), including visitors who are signed in to FlowMic themselves. It also counts against any limit you set on the key. When either runs out, your visitors see that voice input is not available on your site; they are never asked to pay or sign in. If you set up your own engine (§5), that provider bills you directly. Speaking done with your key counts against your plan even if someone copied your key; the per-key limit is how you cap that. If you see use you did not expect, revoke the key in the console and tell us.

Limits and abuse. To protect your allowance and the Service, we limit how often one visitor can start sessions, how long one visitor can speak per recording and per day on your site, and how many people can speak at once through one key. The current values are listed in the website voice input documentation on flowmic.app, and we may change them. We may suspend or revoke a key, or place your account under restricted use (§4), if voice input on your site is used against these Terms.

Your visitors' requests. Pass to us any request a visitor makes about FlowMic's records (github@flowmic.app). Requests about what your website did with the words it received are yours to answer.

10. Contact

github@flowmic.app