Brand Voice: content that reads like you wrote it
The problem with generated content isn't grammar: it's that you can spot it. In Brand Voice you set how you write and the AI applies it to every output; from Pro up, the texts you approve become examples for the next ones.
14 days · 60 minutes · 10 videos · no card
- Tone, audience and vocabulary you set, applied to every output
- AI Memory: the texts you approve become examples for the next ones
- Prompt Library: your own instructions, reusable per output type
- Rewrite in your voice: eleven actions on any pasted text
- A different tone per workspace, on Pro with 1,000 minutes
- Brand Voice already in the trial, the rest with Pro; isolated per workspace, never shared
Brand Voice is the style profile echoflow applies to everything it generates: you set the tone, audience and vocabulary, and the AI uses them instead of neutral textbook prose. Together with AI Memory, the Prompt Library and rewriting, it's what separates correct text from text that sounds like yours.
The real problem with generated content
It isn't grammar: that's been fine for years. It's that you can spot it — same cadence, same transitions, same polite enthusiasm. A post like that on your profile isn't embarrassing, it's simply anonymous, and anonymity is exactly what a brand can't afford.
Four pieces that work together
- Brand Voice — the profile: tone, audience, vocabulary to use and to avoid, a sample text. Set
once, applied to every output.
- AI Memory — the memory: the outputs you approve are kept and reused as examples in later
generations. It's the piece that makes the system improve with use instead of staying at day one.
- Prompt Library — your own instructions per output type, reusable. The place to write "LinkedIn
posts open with a question" once instead of repeating it every time.
- Rewrite in your voice — a standalone tool: paste any text, even text that didn't come from a
video, and choose between eleven actions (brand tone, more formal, more informal, adapt for LinkedIn/Facebook/Instagram/X/newsletter, shorten, bullet list, proofread only).
Isolated per workspace — and that isn't a promise
In an agency running ten clients, the obvious risk is one client's tone leaking into another's copy. It can't happen: Brand Voice, AI Memory and examples are isolated at database level (row-level security), not at interface level. A request trying to read another workspace's data gets nothing back — not even part of it.
The same holds for semantic search and video chat: everything you set or approve stays confined to the workspace it came from.
What we don't do
- We don't clone anyone else's voice. Brand Voice is built on your material: there's no way to say
"write like this author" using someone else's texts.
- We don't guarantee indistinguishability. It helps a great deal, especially with a well-filled
profile, but every output stays a draft to reread before publishing.
- It doesn't pick anything up on its own from what you discard: AI Memory only uses what you
approve as examples. Validate nothing and it stays empty.
- These are text outputs: they don't expire while the account stays active.
Which plan
Brand Voice is included already in the 14-day trial, and of course in the Pro plan: it's the one personalization you fill in yourself, and keeping it out of the trial would leave the «it sounds like AI» criticism unanswered for exactly the people it hits hardest.
AI Memory, Prompt Library and Rewrite in your voice belong to the Pro plan (€19/month for 200 minutes, €49 for 1,000), in both sizes. The rewrite tool draws on the monthly AI tools allowance included in your plan — a single, shared one that resets on the first of the month: it's one of the lighter tools, and roughly, on Pro the allowance is worth around 1,500 rewrites a month. The 1,000-minute size adds separate workspaces, meaning a distinct voice profile per client.
Once the trial is over the account produces nothing new: what you already made — in your voice, if you had set it up — stays readable and downloadable, but generating more takes a plan.
Frequently asked questions
How does the AI write in my tone?
From what you give it. In the Brand Voice profile you set tone, audience, vocabulary, words to avoid and a sample text, and the AI applies them to every generation. From Pro up, AI Memory keeps the outputs you approved and uses them as a reference in later generations. It's a simple, honest mechanism: it follows what you set and approve, it doesn't guess.
Could my tone of voice end up in another echoflow customer's content?
No, and that isn't a marketing promise but a technical constraint. Brand Voice, AI Memory and examples are isolated per workspace at database level: a request trying to read another workspace's data gets nothing back, not even part of it. In an agency with ten clients, one client's tone can't contaminate another's copy.
Can I have a different tone per client?
Yes, on the 1,000-minute Pro plan, which has up to 10 workspaces: each workspace has its own voice profile, its own memory and its own examples. That's what workspaces are for.
What is the Prompt Library for?
For fixing your own instructions per output type — «LinkedIn posts always open with a question», «never use the word X» — so you don't rewrite them each time. They're reusable templates, not a one-off setting you configure and forget.
Does Rewrite in your voice work on text that didn't come from a video?
Yes: paste any text and choose between eleven actions — move it into the brand tone, make it more formal or more informal, adapt it for LinkedIn, Facebook, Instagram, X or a newsletter, shorten it, turn it into bullets, or just proofread it. It draws on the monthly AI tools allowance included in your plan — a single one, shared across every tool: rewriting is one of the lighter ones, and roughly, on Pro the allowance is worth around 1,500 rewrites a month.