Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.
Gladia to join OVH Groupe, accelerating Europe's sovereign voice AI ambitions
PARIS, France — Sept 21, 2026 — Gladia, the AI audio infrastructure provider trusted by more than 300,000 developers and 2,000 enterprise customers worldwide, announced July 31st, that it has been acquired by OVH Groupe, the parent company of OVHcloud and Europe's leading sovereign cloud provider.
Migrating from Azure Speech to Gladia: a step-by-step switching guide
TL;DR: Migrating from Azure Speech to Gladia removes the overhead of custom training pipelines and fragmented per-feature billing. Azure routes diarization, translation, and sentiment through separate services with separate billing meters. We bundle all audio intelligence into one per-hour rate on Starter and Growth plans. Solaria-3 ranks #1 for real-world European business audio, Solaria-1 covers 100+ languages with native code-switching. Both deliver out-of-the-box accuracy that eliminates custom training for most production audio. Most engineering teams complete the API refactoring in under 24 hours.
How to evaluate a speech-to-text API: a technical buyer's framework
TL;DR: Choosing an STT API on vendor benchmarks alone is how teams end up with transcription that looks fine in staging and breaks on production audio. A rigorous evaluation requires building a test set from your own calls, measuring word error rate (WER) on your specific audio distribution, stress-testing latency under concurrent load, and auditing data retraining terms before signing. This guide gives you a reusable engineering blueprint to run that evaluation end-to-end, the same methodology behind our own open async benchmark, which covers 7 datasets and 74+ hours of audio across 8 providers.
GladiaFlow: an open-source voice dictation app for macOS and Windows
Published on Jul 30, 2026
By Ani Ghazaryan
You can explain what you want out loud within a few seconds. Typing the same thing takes minutes, and you'll probably rewrite it once. Most of what any of us writes during the workday isn't prose worth crafting, it's Slack replies, AI prompts, commit messages, meeting follow-up. And for all of it, talking is simply faster than typing.
Today, we’re releasing GladiaFlow – a free, open-source desktop app that turns speech into text in any application. Hold a hotkey, talk, and the transcript appears in whatever field your cursor is in, be it Slack, Notion, Gmail, Cursor, Claude, a terminal, or a form. It runs on macOS and Windows, and is powered by Gladia's real-time streaming speech-to-text API, available under the MIT license.
Here's what it looks like in use:
Why we built another dictation app
This started as an internal itch. We build speech-to-text for a living, and at some point, it seemed silly that we were all still typing. So we put a desktop client on top of our own API and started using it.
It turned out to be good. Good enough that keeping it to ourselves felt like the wrong call. We believe the future is voice, and not everyone who wants to dictate needs a premium tool with a subscription attached. So we're open-sourcing the one we've been using every day.
How GladiaFlow works
GladiaFlow streams audio to Gladia's real-time API while you speak and types the result straight into the focused text field. Hold your hotkey, talk, release. There's no copy-paste step, no separate window to manage, and nowaiting for a file to process after you stop. Transcription happens during the sentence.
That design has a useful side effect: there's no per-app integration to wait for. GladiaFlow types into the operating system, so it works in software that has never heard of it, including internal tools and anything you build yourself. If your keyboard can type into it, GladiaFlow can dictate it.
What you get
Production-grade transcription: GladiaFlow runs on the same real-time API trusted by 300,000+ developers and 2,000+ enterprise teams in production, tuned for real dictation with hesitations, accents, and background noise, not clean studio reads. It has one of the lowest latencies in the market: partials land in under 103 ms and finals in under 300 ms, so words appear as you speak. No waiting for the tool to catch up mid-thought.
100+ languages: Dictate in any of 100+ languages. GladiaFlow covers them all and supports code-switching, so dropping an English term into a French sentence doesn't derail the transcript.
Your vocabulary: Names, acronyms, product names, and industry jargon are the words dictation tools usually mangle. Add your own terms through custom vocabulary and GladiaFlow gets them right instead of guessing phonetically.
Open source and extensible: The code is MIT-licensed on GitHub. Fork it, rebrand it, wire it into your own workflow, or open a PR.
GladiaFlow Specs
Platforms
macOS, Windows
License
MIT (open source)
Languages
100+, including code-switching
Pricing
Free app, pay-as-you-go API usage (~$3/month heavy use)
Latency
Real-time streaming, text appears as you speak
Who's using it
Developers writing prompts: This is the use case that surprised us least and mattered most. Prompts are long, disposable and conversational — the worst possible thing to type and the best possible thing to say out loud. LLMs also absorb imprecision well, so the occasional dictation artifact costs nothing: a dropped comma or a stray "um" doesn't change what Claude or Cursor does with your request. The same goes for the rest of the prose that surrounds code — commit messages, PR descriptions, code review comments, the Linear ticket describing what actually broke, the inline TODO you'll read back in three weeks. None of it compiles. All of it is faster spoken.
Knowledge workers who write in bursts: Nobody sits down for a two-hour writing block anymore; the day is forty small text fields. A Slack reply that needs three sentences of context. The email declining the meeting, politely. A Notion page nobody will read, but someone has to write. Standup updates, retro notes, the follow-up you promised on the call, and are now typing with one hand. Dictation compounds across all of them in a way it never does in a single long document.
Content teams getting a first draft out: Talking an idea through and cleaning it up afterwards is usually faster than staring at an empty page trying to write the first sentence correctly. Outlines, briefs to an agency, the angle for a post you'll rewrite twice anyway: anything where the goal is getting material on the page rather than getting it right. The editing still happens. It just starts from something.
What GladiaFlow costs
The app is free, and transcription is billed through your own Gladia API key on a pay-as-you-go basis. An avid user using dictation for at least half of his daily tasks lands around $3 a month, roughly 4x less than dictation subscriptions charge for the same volume of speech. You pay for what you actually dictate rather than for a seat you forgot to cancel, and if you already build on Gladia, it's the same key you're using in production. Unlock €50 worth of transcription credits, no expiry.
GladiaFlow is open source. If you'd rather read the code, fork it, or run your own build, everything's on GitHub.
That's the whole setup. Enjoy!
FAQs
Is GladiaFlow free?
The GladiaFlow app is free and open source under the MIT license. You pay only for the speech-to-text usage billed through your Gladia API key, which works out to roughly $3 a month for heavy daily dictation.
How does GladiaFlow compare to other dictation apps?
GladiaFlow offers the same core experience as commercial desktop dictation tools such as Wispr Flow — a global hotkey that types into any app on macOS and Windows — with two differences: the code is open source and modifiable, and you pay per use through your own API key instead of a monthly subscription.
Does GladiaFlow work offline?
No. GladiaFlow streams audio to Gladia's real-time API, so it requires an internet connection.
What languages does GladiaFlow support?
GladiaFlow transcribes in 100+ languages with automatic detection and mid-sentence code-switching. It runs on the Gladia API, used in production by 2,000+ enterprise teams, with #1 accuracy in English and robust coverage of other languages.
Can I use GladiaFlow to dictate prompts into Cursor, Claude or ChatGPT?
Yes. GladiaFlow types into the focused text field at the OS level, so it works in any desktop app or browser tab without a dedicated integration. Dictating prompts is the most common way our own team uses it.
Is GladiaFlow available on Linux, iOS or Android?
Not currently. GladiaFlow ships for macOS and Windows only. Since the project is open source, ports are something the community can take on.
Can I use a different speech-to-text provider with GladiaFlow?
Not without modifying the code. GladiaFlow is built specifically around Gladia's real-time streaming API, so switching providers is a fork-level change rather than a configuration setting. Issues, feature requests and pull requests are welcome at GitHub.
Contact us
Your request has been registered
A problem occurred while submitting the form.
Read more
Product News
Gladia to join OVH Groupe, accelerating Europe's sovereign voice AI ambitions
Speech-To-Text
Migrating from Azure Speech to Gladia: a step-by-step switching guide
Speech-To-Text
How to evaluate a speech-to-text API: a technical buyer's framework
From audio to knowledge
Subscribe to receive latest news, product updates and curated AI content.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
