Everything Golem
can do
Voice-to-text that actually understands you. Here's every feature in detail.
Speak.
Golem types.
Golem turns your voice into text and rewrites it however you want. All without leaving the app you're working in. It understands your accent and your expressions.
Transcribes in in a blink
Powered by Whisper large-v3-turbo, Golem processes your voice faster than you can finish your thought. No waiting, no buffering.
Removes filler words automatically
Golem strips out "um", "uh", "like", "you know" and other filler words so your text reads clean. Without you lifting a finger.
“So I was thinking that we should probably move the deadline because the team is really behind”
Select, speak,
transform.
Highlight any text, give Golem a voice instruction, and watch it rewrite instantly. No copy-pasting to ChatGPT.
hey so the quarterly numbers look pretty bad tbh, we should prob talk about it before the meeting tomorrow
The quarterly figures require attention. I suggest we discuss them before tomorrow's meeting.
No more copying to ChatGPT
Stop jumping between apps to rewrite text. Select it, speak your instruction, and Golem transforms it right where you are.
100+ languages.
Your accent included.
From English to Japanese, Portuguese to Korean. Golem understands them all. Speak naturally, get perfect text.
50+ languages supported
Golem supports 100+ languages including English, Spanish, Portuguese, French, German, Italian, Japanese, Korean, Chinese, Arabic, and many more. You can even switch languages mid-sentence.
Switch languages mid-sentence
Start in English, switch to Spanish, throw in some Portuguese. Golem keeps up. AI-powered language detection adapts in real time.
Transcribes in in a blink
Powered by Whisper large-v3-turbo, Golem processes your voice faster than you can finish your thought. No waiting, no buffering.
Make it yours.
Personal dictionary, snippets, and autocorrections. Golem learns your language.
Personal Dictionary
Your words, always right
Train Golem with names, brands, and technical terms so they're always spelled correctly. No more autocorrect battles.
Voice shortcuts for everything
Create voice shortcuts for things you say over and over. From scheduling links to signatures, just speak a cue and get the full formatted text.
Every word, spelled right
Set custom autocorrections so specific words are always written exactly how you want. No more fixing the same typos.
Your data.
Your rules.
End-to-end encrypted
Your audio is encrypted in transit (TLS)
Audio never stored
Processed in real time, then discarded
Your data, yours
We never sell or share your information
Free forever. Unlimited with Pro.
Start for free. Upgrade when you need more.
Free
Pro
Frequently asked questions
Golem Type lets you write by speaking, which is 5× faster than typing. It's for anyone who sends emails, chats on Slack or Teams, takes notes in Notion, writes prompts, does vibe coding, or just wants to save time. If you type daily, Golem saves you hours.
On average, users save up to 20 hours per month. Speaking is 5× faster than typing, and Golem handles formatting, punctuation, and filler word removal automatically.
No. Your audio is processed in real time and immediately discarded. We never store recordings or transcripts. Everything happens over encrypted connections (TLS).
Your account data is stored in Supabase with Row Level Security (RLS) enabled. Only you can access your own data. Audio is never stored. All API communication is encrypted end-to-end.
Golem supports 100+ languages including English, Spanish, Portuguese, French, German, Italian, Japanese, Korean, Chinese, Arabic, and many more. You can even switch languages mid-sentence.
Get 5 hours back every week
Join thousands who speak instead of type. Free forever with 2,000 words/week.
Get Started, Free