FluidTalk - AI Chat Funnel Engine for Creators & Agencies | AI Characters for DMs + Comments on 10+ Platforms | $0 Plan Available

If you run chat for creators, an OFM agency, or any DM-driven offer, you already know where the money leaks: replies that take hours, leads nobody follows up with, chatters who burn out on night shifts, and a different "voice" on every platform.

FluidTalk replaces the typing, not the strategy. You design the persona, the knowledge and the funnel. The AI Character runs it, around the clock, across every account you connect.

• A cloud engine + dashboard that generates the messages your AI Characters send: first-touch cold DMs, replies, public comments and follow-ups - and tracks where every lead sits in your funnel.
• It does not log into your accounts and does not post anything by itself. Messages are delivered by a connector of your choice: FluidTalk Extension (browser - extension.fluidvip.com), Onimator (Android - onimator.com), Xmator (X, cloud - xmator.com), ReddMator (Reddit, cloud - reddmator.com), or your own bot through the FluidTalk API.
• Usage-based: what you pay for a plan becomes wallet credit, and it is spent per generated message (full pricing below). Recordings of it running on Instagram, Telegram and WhatsApp are further down this post.

---

Ready to run on: Instagram - Threads - WhatsApp - Telegram - TikTok - Reddit - X (Twitter) - Snapchat - Tinder - Bumble

The platform list is not a hard limit. A platform is just a name you bind a Character to, and the messages are delivered by whatever connector you point at the API - so if you can read and send messages on a site, you can write a connector for it and run a Character there. Same with the funnel: you decide the hops. Platform to platform, platform to your own site, landing page, shop or booking form - you draw the route and the Character walks leads down it.

---

1. AI Characters - a persona you fully control

• Identity: name, age, origin (the timezone is derived from it, so the Character keeps a realistic daily rhythm), job, bio, personality, and a private conversion goal the lead never sees.
• Voice: lowercase or not, emoji density, personal slang, and example lines written in the Character's own words.
• Self-facts: the list of things the Character is allowed to share about itself.
• Languages: a default language, or mirroring whatever language the lead writes in.
• Chat bubbles: replies arrive as 2-5 short messages with human-like delays between them, instead of one wall of text.

2. Lorebook - the Character's knowledge base

• One entry per fact: your pricing, what you sell, house rules, the story behind a content set.
• Three triggers per entry: constant (known on every turn), keyword (fires on words like "price" or "how much"), or semantic (fires whenever the topic comes up, however it is worded).
• Priorities decide which facts win when more of them match than fit in one turn.

3. Lead memory & profiling

• Profile probes: tell the Character what to find out about each lead (job, city, budget, what they are into). Answers are stored and reused later in the conversation.
• One lead, every surface: the same person commenting on your post and messaging you on another platform is recognised as one lead with one shared memory.
• Multi-account control: one shared conversation per lead across all your accounts, or a separate session per account. Optional one-reply-per-lead de-duplication across accounts on paid plans.

4. Visual strategy engine (workflows)

• Build the funnel as a graph: phases (for example Cold - Warm - Pitch - Converted / Re-warm) connected by rapport gates and conversion gates.
• Each phase holds a deck of cards that unlocks specific moves: banter, mirroring, asking about the lead, sharing about itself, callbacks to earlier messages, soft pitch, link drop, objection handling and more.
• Tune the pace: how many messages a typical lead takes before the link, minimum turns before a pitch, soft-pitch and hard-drop thresholds, climb speed, and how many re-warm attempts follow a missed conversion.
• Write your own moves and cards, re-word any move per phase, and route objections to your own replies.
• A validation report flags structural problems - such as a phase no lead can ever reach - before you put a workflow live.

5. Cross-platform funnel routing

• Map the hops your leads take (for example Instagram - Telegram - your landing page) and the link the Character may share at each step.
• Per hop, choose auto (it moves the lead when the moment fits) or offer (it asks first).
• Handoff moves ask for the lead's username on the next platform, so your team or your next account can pick the conversation up there.

6. Entry points - how conversations start

• Inbound chat (the lead messages first) or cold outreach (the Character opens).
• Event triggers with their own openers: story reply, reaction or mention, poll vote, quiz answer, close-friends add, new follower, keyword comment, keyword DM, new match, and more.
• Build your own entry-point packs with custom payload fields for any event your connector can detect.

7. Automated follow-ups

• The Character writes re-engagement messages for leads who went quiet - one directive per follow-up step, written by you or taken from the built-in defaults.
• Built-in limits: one per lead per day, a minimum gap between them, a 90-day cap, and silence around a pitch.
• Queue, send through your connector, acknowledge - so nothing goes out twice.

8. Public comment engagement

• The Character reads the post - caption and images - and writes a specific comment in its own voice.
• Full-thread replies or comment-once mode, reply caps per thread and per person, and a reply-to-strangers toggle.
• Drive-to-DM: once a commenter reaches the warmth threshold you set, the Character invites them into DMs - with no link posted in public.
• It bows out of hostile threads instead of arguing in public, and warmed-up commenters become leads your DM funnel already knows.

9. Media vault

• Upload your images into the Character's own hosted vault.
• Vision analysis captions each image (location, outfit, activity, mood, framing), so a request like "send one from the beach" resolves to the right file.
• Sending rules: the content level a conversation has earned, the time of day on the Character's local clock, and shoot-day sets that keep the look consistent within one day.
• A selection tester shows which image would be sent for a given moment - and which rule blocks the rest.

10. Lead filtering (Agency plan and above)

• Mute leads from selected countries, or from lower-income tier countries, so your budget goes to conversations worth having.

11. Templates & marketplace

• Clone ready-made Characters that arrive with a persona and a working conversation flow - inbound, keyword, cold outreach (X, Reddit), B2B and no-flirt styles.
• A reusable library of moves, cards and decks you can share across all your Characters.
• Publish your own Characters, workflows and entry packs (each one is reviewed before it is listed).

12. API, SDKs & AI agents

• A per-Character connector token plus the Characters API: push an inbound message, get the reply bubbles back.
• Official Node.js and Python SDKs with typed responses.
• MCP server: connect Claude, Cursor or an OpenAI Agents app, and let your own AI agent create, configure and test Characters for you.
• Test before you go live: sandbox chats, comment previews, workflow validation, and a conversation view showing each lead's phase, rapport and whether the pitch has been made.

---

Short, unedited screen recordings, looping at real speed. In each one the FluidTalk Extension panel sits on the right: it reads the thread, generates the reply bubbles, and either inserts them for you or sends them on auto-reply.

Rapport, the link drop when the gate opens, and the lead confirming he subscribed - with the panel showing replies, requests and comment replies as they happen.

Opening a stranger from his profile, with the Character writing the first line off his feed.

The same Character on a second platform, including a queued follow-up for a lead who had gone quiet for 26 hours.

Inbox view with auto-reply on, and the follow-up counter working through a thread that had stalled.

17 Android devices running Instagram DMs at the same time. FluidTalk writes, Onimator delivers - the same setup works with the Extension, Xmator, ReddMator or your own bot on the API.

---

Your subscription is your conversation budget - there is no separate usage bill on top of it. Every dollar you pay for a plan lands in your wallet as tokens (1 token = $1) and is spent on conversations at your plan's rate. A bigger plan simply buys more conversations for the same dollar.

Usage is metered per generated message. Every message a Character writes costs a small fraction of a token, so a lead who closes in ten messages costs less than one who chats for a week - there is no flat fee per conversation.

The figures below are given per conversation only so you can compare plans: they assume an average conversation of up to 50 sent messages. Your real cost per lead lands wherever your conversations actually end.

If your plan lapses, the tokens you have not spent stay in your wallet and you keep using them at the rate of whatever plan you are on. A message the wallet cannot cover is refused before it is generated, so there is no overage bill to be surprised by.

• Yearly billing: 12 months for the price of 10.
• Both conversation columns are estimates built on that ~50-message average, because billing happens per message - short chats stretch the budget further, long ones eat into it. The exact cost of every message comes back with each turn, so nothing is hidden.
• Plans renew automatically each billing period until you cancel. You can cancel at any time in the dashboard, and the plan stays active until the end of the period you paid for.
• Connectors (FluidTalk Extension, Onimator, Xmator, ReddMator) are separate tools with their own pricing - or run your own bot on the API at no extra cost.

---

• Account and plan: activated instantly after payment.
• First working Character: as soon as your persona is set up, it is ready to chat. Cloning a template takes minutes; a persona written from scratch takes longer.
• Going live on a platform: same day with the FluidTalk Extension, Onimator, Xmator or ReddMator; longer if your developer builds a custom bot on the API.
• Support replies: within 24 hours on Telegram, Discord or BHW DM.

---

Good fit: OFM and creator agencies, chatter teams, social media marketers running DM funnels - anyone with more inbound DMs or comments than they can answer quickly.
Not a fit yet: if you barely get messages, FluidTalk will not fix that. It works on the traffic you already have; it does not bring you followers or subscribers by itself. Fix distribution first, then come back.

---

• FluidTalk is an AI and automation based tool. Every message is AI-generated from the persona, knowledge and workflow you configure.
• Automated messaging can be against the terms of service of some platforms, and account restrictions are possible. Pacing controls, bubble delays, reply caps and the no-links-in-public-comments rule are there to lower that risk, not to remove it.
• Delivery depends on third-party connectors or your own bot. The FluidTalk engine itself does not ask for your platform logins.
• You are responsible for how you use your Characters, including any AI-disclosure and consumer rules that apply in your country and on the platforms you run on.

---

Is this just a ChatGPT wrapper?
No. A Character combines a persona, a triggered knowledge base, per-lead memory across platforms, and a phase-based workflow that decides what it does on every turn. Pitch timing, follow-ups, handoffs and comment behaviour are all rules you control.

Do I need to code?
No, if you use one of the ready connectors (FluidTalk Extension, Onimator, Xmator, ReddMator). If you want your own bot, the API is a single endpoint with Node.js and Python SDKs.

Can I try it before paying for a plan?
Yes. The $0 plan gives you 1 Character, and you only pay for the conversations you actually run.

Can one Character run several accounts?
Yes. A Character can be bound to several platforms and accounts, with shared or separate lead sessions per account.

Which languages does it support?
Any default language you set, or automatic mirroring of the language the lead writes in.

Can it send images?
Yes, from your media vault, following the stage, time-of-day and set rules you configure.

What happens when my wallet runs out?
New replies are refused before they are generated, so there is no overage. Top up or renew your plan and your Characters carry on - and any tokens left over stay in the wallet.

Do you help with setup?
Yes. On the Scale and Enterprise plans we walk you through your first Character and connector. Ask on Telegram or by BHW DM.

---

5 x review copies: 1 month of the Agency plan plus $20 of wallet credit.
Who qualifies: Jr. VIP members with 100+ posts who run DM funnels today.
How to apply: reply in this thread with a short description of your setup, then send me a BHW DM.
Review deadline: an honest review in this thread within 14 days of activation.

---

Card (credit / debit) and crypto. No PayPal Friends & Family, no Western Union.

---

You can build a Character and run it in the built-in Simulator on the $0 plan before you spend anything, so you can see exactly what you are buying first. For that reason conversations that have already been generated are not refundable - that work has been delivered. Here is what we do refund:

• Unused wallet balance: whether it came from a plan payment or a top-up, it is refundable on request within 14 days of that payment, minus the tokens already spent.
• Not as described: if a feature listed in this thread does not work on your account and we cannot fix it within 7 days of your report, you get a full refund of the current plan period plus your unused balance.
• Service outage: if the FluidTalk API is unavailable for more than 24 consecutive hours, you get a pro-rata refund or account credit for the affected days - your choice.
• Duplicate or accidental charge: refunded in full.
• Not covered: tokens already spent on generated conversations, restrictions or bans applied by third-party platforms, and fees for third-party connectors, which are not our products.
• How to request one: send me a BHW DM or message us on Telegram with your account email and the transaction date. Approved refunds go back to the original payment method within 5 business days.

---

1. Create your account and pick a plan: click here to start with FluidTalk
2. Top up your wallet in the dashboard.
3. Clone a template or build your first Character, bind a platform, connect your connector - and go live.

---

Telegram community: https://t.me/+7TmYLy0KKqZiZDE0
Telegram updates channel: https://t.me/fluidvip
Discord: https://discord.gg/PXUeVtPV6
On BHW: DM me

Summary
This service has been tested by the Marketplace team to ensure that it meets the standards for the Marketplace. The service has been checked to ensure that it can deliver what it advertised in accordance to its TAT or in a timely manner if no TAT was provided.
Service Information

• Disclaimer: We've allowed the seller to select a single person and give their top-level package for free. The staff monitored the process and everything involved with it.
• Here are a few of the lines from the review left by the member:
  • "The conversations felt natural, memory worked properly, the Characters stayed consistent with their personas, and language mirroring worked well across the languages I tested."
  • "Overall, my experience was positive. The core features worked well, and the few issues I ran into were mostly things that could be adjusted in the configuration."

Thread Edit Log

• The 5 most recent thread edits will appear here

Spoiler: Thread Edit Archive

With anything more than 5 going in here

Information For Buyers:

• Service Quality: The quality of the product or service that we receive is what you should expect, or better. If you feel that the quality of the service has dropped significantly, please let us know via the report button.
• Disputes: If you do not receive a product or service as advertised, or at all in the event of a dispute, do not be afraid of a "no refunds" refund policy as you are allowed to request a refund through the Dispute Resolution process. For more information and to see whether or not your dispute qualifies, refer to the dispute rules and procedures.
  IMPORTANT: For full clarity, taking the transaction off BlackHatWorld removes your ability to run a dispute - we only accept disputes for transactions run through the forum with traceable evidence (e.g., DM between the seller and the buyer, leaving a comment in the sales thread that you made a purchase, etc)
• Review Copies: Per the Marketplace rules, if a seller does not offer trials or review copies, please do not request them in the sales thread, otherwise your post will be removed and further action may be taken.

Quick bump with something worth knowing if you were on the fence: the fastest way to see what a Character actually does is the FluidTalk Extension - a browser extension you run on accounts you already use. No server, no bot to write.

What the Extension does

• Sits as a panel next to your inbox on Instagram, Telegram, WhatsApp and more. It reads the open thread and generates the reply bubbles from the persona, knowledge and workflow you configured.
• Two ways to run it: insert mode (it drafts, you read and hit send) or auto-reply (it sends for you, with human-like delays between bubbles).
• Install it, bind a Character to the account, and you are live - the same reply bubbles you see in the recordings in the first post.
• Pairs with the $0 plan: build one Character and only pay for the conversations you actually run. Full pricing and the other connectors are in the first post above.

---

The Extension is just one way to deliver messages. If you run your own bot or software, the Characters API gives you the same engine behind a documented endpoint, with docs deep enough to build against from day one:

• Push an inbound message, get the reply bubbles back - one per-Character connector token, one endpoint.
• Official Node.js and Python SDKs with typed responses, plus documentation covering entry points, workflows and the conversation state you read back per lead.
• MCP server so Claude, Cursor or an OpenAI Agents app can create, configure and test Characters for you.
• Test before you ship: sandbox chats, comment previews and workflow validation, so you see the behaviour before a single real DM goes out.

Whichever route you take - extension or your own integration - the Character, the funnel and the lead memory are identical. Questions on either one? Send me a BHW DM and I'll point you at the right docs.