Seed: Voice Brain Dump

Get it off your mind. A thought, a reminder, a frustration or something you are still working through. Say it to Seed and carry on with your day. You do not need to organize your thoughts before you speak. SAY WHATEVER IS ON YOUR MIND Press once and talk on iPhone or Apple Watch. A few words between tasks or a longer thought on a walk: there is no outline to fill in. Watch recordings go to your iPhone for transcription. You can also start with Siri, Shortcuts or a supported Action button. COME BACK WHEN YOU HAVE ROOM Timeline brings your spoken notes together by day and time. Read back, search for a phrase or highlight a passage worth returning to. Turn selected paragraphs into bullets or checklists yourself. Some thoughts may become a paragraph, a decision or an action. Others have served their purpose just by being spoken. You decide what to make of them. With Apple Intelligence available and enabled on compatible hardware, Seed makes a cleaner reading version and short titles for new recordings. Your original words and audio remain available with each note. Listen at your preferred speed while the words follow along. ROOM TO SPEAK FREELY Audio transcription runs on your iPhone. No account, no subscription and no data collected by us. Reading versions run on the device by default. Eligible iOS 27 devices can optionally use Apple Private Cloud Compute for text cleanup and titles, subject to Apple's usage limits. An optional webhook sends original transcripts to your own tools. KEEP WHAT YOU WANT TO RETURN TO New installs move notes older than seven days to Recently Deleted. Choose 1, 7 or 30 days, or Never, in Settings, Recordings, Auto-delete. Existing choices are preserved. Deleted recordings can be recovered for 30 days. Choose Never if you want to keep notes until you delete them. WHEN YOU WANT MORE CONTROL Use Apple Speech with a chosen language, or download Everyday for automatic detection across 25 languages. Compact offers a smaller download and transcribes while Seed is open. Downloads retain progress when paused; unused optional models can be removed. Vocabulary spelling pairs and optional filler-word removal apply to reading versions. High voice sensitivity can separate quieter background speech in the reading version without changing your original. Available in English, Spanish and German.