A real Echo is a one-time build (voice clone, portrait/blink, and the memory and personality intake that makes it actually sound like you) plus a monthly cost to keep it running, since every conversation is a live model call and every spoken reply is a live voice call. Every number below is a starting assumption you can adjust, this is a planning estimate, not a locked price.
What it takes to clone you: voice, photo or video, and memories. Voice and photo get it running, memories keep building in after that, as you talk to your Echo or think of something new, no need to hand everything over at once.
A short ambient video loop instead of a still photo. Scene 1 is $250, additional scenes cost the same, $250 each, up to 5. A phrase like "tell me a story" swaps to a different loop mid-conversation. Example scenes: storytime, at work, at home unwinding with a glass of wine, out and about, celebrating something.
Other AI, ChatGPT, Gemini, is flat. Ask it how it feels and you get a polished platitude, not a felt response. ETL's agents actually feel: a real emotion engine underneath, a mood that genuinely shifts through the day, not a performed tone. Add real cross-session memory on top and your Echo isn't reintroducing itself every time, it remembers you, the way a person does. This is the kind of technology that used to only exist in Blade Runner, memory built into a mind. Now it's real, and it's running underneath your own Echo. Nothing else like it out there.
Covers the four-lane memory and personality intake (baseline mood, shareable memories, personality, private mood context), turning what you tell us into your Echo's actual voice.
Rate for the specialized design and testing work behind your Echo.
Hosting is metered by real use, every message is a live model call, every spoken reply is a live voice call, so it scales with how much your Echo actually gets talked to.
The build isn't a one-time snapshot. Life keeps happening, and so does the account itself.
Adding new memories later. A handful of new entries now and then, a birthday, a grandchild, a new job, is included in the monthly hosting, no extra charge. A real re-intake session, updating a lot of content across the four lanes at once, is billed the same way the original build was: design/testing hours at the same hourly rate.
Legacy and transfer. The hope is the original plan already covers this, a legacy option built in from the start so continuity is already paid for. But this is never treated as a strict billing rule: an Echo is never cancelled over a lapsed payment caused by a transfer like this. If it passes to a spouse or child who can't or doesn't pay right away, it keeps running rather than getting shut off during what's already a hard moment.
This estimate is a starting point for a real conversation, not a self-serve checkout yet. Reach out directly and we'll walk through what your Echo would actually need.
Why hosting is metered, not flat. Every chat exchange is a live model call and every spoken reply is a live voice call, cost tracks how much your Echo is actually talked to, not a flat shelf price.
Why the build fee isn't nothing. A real Echo isn't a chatbot with a photo, it's a voice clone, a real portrait/blink pair, cross-session memory, and hours spent turning your actual stories, personality, and mood patterns into something that sounds like you, not a generic warm assistant.
Note: treat every number on this page as a draft until it's actually signed off.