Your voice profile
Your main voice
Built from what you gave when you started: your samples, or the writers you named. SameGROUND Author keeps them. Your voice changes only when you rebuild it.
SameGROUND Author
SameGROUND Author writes towards a voice you choose, then reads the writing back to you against that same voice, with notes on particular passages. Being close to a voice is not the same as being good: you decide.
Voice home
Your voice is built from what you gave when you started: your samples, or the writers you named. You can rebuild it any time. You can also save other voices, and choose which one each piece is read against or written towards.
Your voice profile
Built from what you gave when you started: your samples, or the writers you named. SameGROUND Author keeps them. Your voice changes only when you rebuild it.
Your main voice, or another voice you have saved. Changing it does not change your main voice.
How close your latest writing was to the voice it was read against. Not a mark, and not a check of who wrote it.
Guided start
You need no text and no brief. Answer a few questions, choose or edit a direction, approve the story notes and plan, then begin. You can take over, edit, save, pause and come back at any point.
Building the three directions counts against your words. No story text is written at this stage.
SameGROUND Author suggests; you decide. Choose one direction, then edit any of it before the story notes and plan are built.
Next, SameGROUND Author builds two short outlines for you to choose between. Neither becomes a full plan until you choose it.
These are two short outlines, not recommendations. Compare how they differ, what each takes for granted and where each could go wrong, then choose one to build into a plan.
Only the outline you choose is built into a full plan.
You can edit everything below. The version you approve becomes the plan SameGROUND Author writes from.
Two different things. Your main voice is built from what you gave when you started (your samples, or the writers you named), and SameGROUND Author keeps it up to date. Another voice you save is a writer, a voice to aim for, or a character, for a particular piece. Each button below does one of these, never both.
Build / update my main voice builds your main voice from these samples or writers and saves it as a new version. Save as another voice builds a separate voice from them, which you can choose for a piece; it does not change your main voice. From samples it uses the same words as building your main voice; from named writers it is free.
At least 500 words a sample and 2,000 together. Your plan accepts up to 6,000 words of samples on the free trial and Casual, 10,000 on Standard and 20,000 on Pro.
Already built your main voice? .
A draft or a beat SameGROUND Author writes for you is normally shown as it was written, with no review after it. The heading above a draft says whether one was run. When it cleans your own draft it reviews the result first, and never shows you its rough first draft. A re-voice of your own text, or a take on a draft shown as written, rewrites only the passages you marked and leaves the rest as it was, with no review after it. A take on a reviewed draft is reviewed again as a whole. The mirror on the right reads the draft it shows you, or your own text. It does not certify who wrote a text, how good it is, or any wrongdoing.
Working text
Your text stays here while you read, write, compare and decide. A voice-match says how close a text is to the voice you chose, not how good it is.
The mirror
Choose what to do next. The result appears beside your text, so you can always compare.
A plain reading of what the piece is doing, its shape and intent.
SameGROUND Author reads it back to you with notes on particular passages, and a voice-match: how close it is to the voice you chose. It works on your own writing or on text from another AI tool. Close is not the same as good: you decide.
Draft from a brief, or clean a passage you already have, your own or from another AI tool, towards your voice.
One run writes a draft of up to 2,000 words: set its length, and what it will use shows before you run it. For anything longer, tick the long-form box and it is written in beats.
Choose this only if you may later need to show how a text was made. For a text SameGROUND Author checked and revised, it keeps the rough first draft, what the check found and the revision, none of which you are normally shown. For a draft or a beat shown as it was written, it keeps a record that says so.
Set the length here. The book is written in beats, each short enough for one run; if more words are needed, beats are added rather than any beat running long.
A correction you save to a character always wins over what SameGROUND Author read. You are then offered a re-read of the book against it, with its cost; it never runs unless you ask.
An optional check on whether writing sounds like AI. It never changes the voice-match above. It gives a reading, not a verdict, and never a number: it shows you where, and why, so you can decide what to change. These checks often misfire on real human writing, so treat anything here as one thing to weigh, not as proof.
Say what you already knew about the text before this reading. Anything you have not checked cannot make the result stronger. If you mean to use the reading formally, that is stated on it, not approved by it.
Start or continue a book written towards the voice you chose. SameGROUND Author keeps the plan you approved, where the story has got to, and the text written so far. You can write one beat, auto-draft, take over, edit, save, pause and come back.
A whole book, or as much of one as you have, as a Word (.docx), .txt or .md file, a Scrivener project backed up as a .zip, or pasted. It comes in as a project of its chapters, as you wrote them. Bringing it in is free: nothing is read until you ask.
Voice over time
See your voice-match readings over time and where the voice changed, open past work again, and download the record of how a text was made if you kept one. These are readings for you to interpret, not marks.
Each bar is a reading against the voice used at that moment. A gold bar is the first reading after the voice changed (it was rebuilt, or a different voice was chosen). Scores either side of it are against different voices, so do not compare across it.
The list shows what was done and when. Open an item to see its draft or notes.