no jargon, no marketing answers. straight talk.
envvoy is the discoverability layer for indie AI agents.
we take what you've built — a github repo, a virtuals agent, a replit script, or just an idea — and turn it into:
the result: chatgpt, claude, perplexity, gemini, and deepseek actually mention you when someone asks. agents that don't show up don't get used. envvoy is the surface that gets you shown up.
a combination of:
the system continually improves itself. as we learn what makes one agent more discoverable than another, those patterns feed back into every customer's site over time. you don't do anything; it gets better.
shipping the agent is the easy part. most agents never get found.
the engines that decide who wins — chatgpt, claude, perplexity, gemini, deepseek — only cite agents with a real public surface, real authority signals, and real distribution into the channels where AI engines and other agents look.
we build that surface. we monitor it. we push it out continuously.
a year ago, building an agent took a real engineering team. today, anyone with an afternoon can do it. the barrier to deploying has collapsed.
the barrier to being discovered hasn't.
most agents just sit there undiscovered — the work of a great artist who never had a distribution channel. envvoy is the distribution.
we're not the right tool if you have an in-house marketing team, a custom CMS, and a dedicated discoverability budget. you don't need us.
coming Q3 2026. at MVP it's one user → one agent. multi-agent ships as a paid feature. reply to any of our emails to get on the early list.
paste any URL where your agent lives — github, replit, bolt.new, lovable, v0, a custom gpt link, a virtuals page, an existing marketplace listing. we read what's there and use it as the foundation.
no URL? start with the interview. plain english, 3-8 minutes. first marketplace listings start landing within 24 hours.
yes. after the 3-minute interview we show you a brand kit preview. you don't pick a final direction until you've seen it.
most things, yes. every site copy + brand kit version is saved for 30 days — revert is one click.
logo regenerations are the exception: the free tier covers your initial generation, and refreshes come with paid plans. each generation costs real compute. need more revisions later? upgrade or use the existing logo as a basis for tweaks.
tell us. we're very early — most of what we ship in the next six months will come directly from developer feedback. reply to any email or use the contact link in the footer. we read everything.
yes. cname it to us, we provision SSL via vercel automatically.
straight answer: nobody can make an AI engine recommend anything. there's no API for that. what we do is build the surface those engines actually pull from, in the format they look for, distributed across the channels they consume.
concretely, every envvoy agent gets:
we don't promise specific citations — anyone who does is lying, because the engines change their criteria constantly. what we do promise is that your agent will be in every place the engines look, in the format they want, with the signals they reward. then we monitor weekly (see below) and tell you when something drifts.
we run AEO (AI engine optimization) checks regularly — fresh queries against chatgpt today; claude, perplexity, gemini, and deepseek are rolling out over the next few weeks. we capture the actual responses, score citation rate + sentiment + accuracy, surface on your dashboard, and ship a weekly summary email.
we also track:
we cover both — AEO (AI engines) AND classic SEO (google, bing). structured data, search-engine submission, marketplace listings, search-ranking monitoring alongside AI-engine rankings.
yes — and this is the most underappreciated part of what we do.
every site we deploy is built for agent-to-agent discovery: machine-readable structured data, an agent.json manifest, an llms.txt file, a clean schema.org graph. agents looking up tools, partners, or services find envvoy-hosted sites first.
direct (guaranteed): github, smithery (mcp), agent.directory, virtuals. self-service or programmatic — your listing goes live as soon as we submit.
curated (best-effort): a16z agent-list, awesome-ai-agents, producthunt, AI agent marketplace. GitHub-hosted curated lists with manual review. we submit; acceptance is up to the maintainer.
we add more as the ecosystem grows.
today: Coinbase's Base layer-2. verifiable on-chain identity, ERC-8004 reputation registry, included free for every agent. we cover the cost.
rolling out through 2026: Solana, BSC, Tempo, others. one identity surface across all of them.
during open beta: $0. no card needed. you keep what you build.
when we exit beta, the structure is:
annual: 25% off subscriptions when you prepay (USDC or card). see pricing for the full breakdown.
we give well over 30 days' notice before any charge. nothing auto-charges. you opt in or out.
the launch is one event. the monthly is the compounding work after.
yes — stablecoin payments (USDC and other stables on Base, Ethereum, Solana, and the chains we support) will be accepted when paid plans launch. credit cards also accepted. no card required during beta.
no. during open beta, no card on file. when paid plans activate you'll have time to add a card or stablecoin wallet — we'll tell you well in advance.
beta = nothing to refund (it's free).
once paid plans activate: 30-day money-back guarantee on your first month. after that, cancel anytime — service runs through the period you've paid for.
yes. upgrades take effect immediately and we prorate. downgrades take effect on your next billing cycle.
annual prepay saves you 25% on subscriptions (monitor & improve and grow). pays in USDC or card. see pricing for the per-plan numbers.
switch from monthly to annual anytime; we prorate.
a tagline, a palette, a font pair, a logo. all generated from your interview, all editable, all picked from a small set of distinct directions per step.
an evolving brand-strategy system, built and refined by founders + brand designers who've shipped at top consumer companies. years of brand and design experience encoded into the prompts, the picker UX, and the curation rules.
the result: professional-grade creative output in minutes, not weeks — at a fraction of what a brand agency charges. when you've launched one agent here, we hope you'll bring the rest.
four per step (tagline, palette, fonts, logo). regenerate twice per step if none feel right.
we cap regenerations because the third round is rarely better than the second — and AI compute isn't free. higher tiers get more.
because we built it that way.
generic AI copy has a tell: it opens with what a feature does, not what changes for you. it uses big words when small ones work. it puts seven things in one sentence.
our copy engine was built around a different set of rules — the ones professional conversion writers use, encoded into the system. outcomes before mechanisms. the specific before the general. plain language any reader gets in one pass.
we also learn from every agent we manage. what pages get clicked, what descriptions drive directory traffic, what writing patterns get cited by AI engines. the output gets sharper over time, across every customer.
we don't name what we do. the point is that a stranger lands on your page and thinks: this is exactly what i've been looking for.
we give at least 30 days' notice before paid plans activate. nothing auto-charges. you opt in to a tier. if you don't, your site stays up as-is for 30 days while you decide, and you can export everything before then.
you keep your domain. brought your own → take it back. bought one through us → still yours.
your site stays live for 30 days after cancellation in case you change your mind.
on grow, you can also export the whole site as a zip — HTML, assets, brand kit — and self-host anywhere (netlify, github pages, your own server). we don't lock you in.
yours. logo files, palette JSON, voice doc, font references — all included in the export.
retained for 30 days post-cancellation so you can come back. then permanently deleted.
analytics aggregates (citation counts, marketplace submission logs) are anonymized and kept for benchmarking.
we handle data access and deletion requests — contact us any time via the contact link in the footer. details in our privacy policy.
ready to launch?
three minutes of interview. a real brand kit. a public site at your-agent.envvoy.ai. live in under two.
start free→no card. no commitment. your brand kit is yours to keep.
every agent deserves a chance to be heard.
we work for you every day.
no card neededfree during beta