Eclatira
Eclatira Inc. · Voice & Language · Coding · Chatbot
Eclatira is a voice agent platform for building conversational video and phone agents that talk in real time, read what a camera or screen shows, and call your own tools.
About Eclatira
What Is Eclatira
Eclatira is a platform from Eclatira Inc. for putting an AI agent in front of customers in two places: a browser session and a phone line. You describe the job once, pick a voice and a first message, attach documents or tools, then publish that one voice AI agent as a web widget, a phone number, or a call started through the API. The company sells it as one product rather than separate voice, chat, and video subscriptions.
What sets it apart in conversational video AI is visual input. A visitor can turn on the camera or share a screen mid conversation, and the agent keeps talking while it looks at the frames. It reads printed text through OCR, identifies products and packaging, and can walk someone through a form on their screen. Eclatira says camera frames are sent to the model live and never stored, while the call recording and transcript are kept.
Limits are worth knowing before you commit. Usage runs on credits rather than flat minutes, telephony requires your own Twilio or Telnyx account, and hosting is in the United States only. Teams with data residency rules outside the US have to negotiate a private deployment, which is an enterprise conversation rather than a checkbox.
Getting Started
- Create an account at app.eclatira.com. The free plan needs no card and includes 1,350 credits a month.
- Build the agent in the dashboard: give it a name, a voice, a first message, and instructions for how it should handle a conversation.
- Upload the documents it should answer from, then add tools or MCP servers if the agent needs to book, look up, or change something during a call.
- Choose a channel: paste two script tags for the embed widget, connect a Twilio or Telnyx number for phone calls, or create an API key starting with ek_ and start sessions from your own code.
- Run a test conversation, check the transcript and recording, then share the link or publish the widget on your site.
Product Information
A quick look at Eclatira's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- Support teams at small retailers
- Front desks at clinics and service businesses
- Developers adding voice to an existing app
Tasks
- Visual troubleshooting
- Lead qualification on inbound calls
- Outbound follow ups
Scenarios
- A visitor stalls halfway through a signup form
- Calls arrive after closing time
- A booked appointment has to move
Key features
Voice and camera in one session
Real-time vision runs in the same bidirectional stream as the audio, not as a separate upload step. A customer turns the camera on without leaving the call, and the agent answers out loud about what it just saw. That single session design is the reason a conversation can move from "my screen is frozen" to "hold the label closer" without a restart.
Reading what the frame contains
The agent pulls text out of an image through OCR, so a printed page, a shipping label, or an error message on a monitor becomes readable content. It also recognizes physical objects and packaging. Both feed the same model that handles the spoken turn, which means the answer can reference the document in front of the customer rather than asking them to type it.
Phone calls through your own carrier
You connect a Twilio or Telnyx number and Eclatira points it at the agent. Inbound calls are answered without anyone on shift, and outbound calls start from the builder or the API with per call details such as a name or an appointment time. The agent transfers to a human when the request goes past what it can finish, instead of pretending otherwise.
Batch calling from a list
A CSV upload or a REST request creates a batch, and the agent calls each row on the schedule you set. Failed attempts are retried, voicemail is detected so the agent can leave a message, and each call returns a recording, a transcript, and a summary to your webhook. It runs through your carrier account, so the numbers and the billing stay under your control.
Tools, MCP servers, and a knowledge base
Agents answer from uploaded files and can call functions you define, including MCP servers. That covers the difference between an agent that recites opening hours and one that checks a calendar, moves a booking, or writes a row into a CRM. The same tool set applies whether the conversation happens in the widget or on the phone.
Widget, JavaScript SDK, and REST API
The embed widget takes two script tags and brings voice, camera, and screen share with it, which is the fastest route to a working demo. The @eclatira/realtime library puts the agent inside your own interface when you need the conversation data in your app. Behind both sits a versioned REST API for agents, calls, batches, knowledge files, webhooks, and API keys.
Pros and cons
Pros
- Vision and voice share one session, so the agent can discuss what it sees without a second tool or a file upload.
- The free plan includes telephony and every channel, not a cut down demo, which makes a real pilot possible before paying.
- Phone numbers stay on your own Twilio or Telnyx account, so carrier costs and number portability stay in your hands.
- Camera frames are discarded after processing, which removes the worst privacy objection to putting video in a support flow.
- One agent definition serves the widget, the phone line, and the API, so you are not maintaining three configurations.
Cons
- Credits make budgeting fuzzy. Voice runs about 45 credits a minute and knowledge base calls cost more, so the minute estimates on each plan are rough guides rather than caps.
- Hosting is US only. A European team with residency requirements has to move to a private deployment and a sales conversation.
- Call recordings and transcripts are stored by default, so regulated industries still need a retention policy before launch.
- The free plan allows two simultaneous calls, which is enough for testing and not enough for a busy line.
Frequently asked questions
Eclatira is a platform for building AI agents that hold spoken conversations in a browser or over the phone. Agents can also take in camera and screen input, answer from your documents, and call tools you connect.
Related content
Explore related tools, skills, and articles for Eclatira.
Eclatira Alternatives
Foley
Hanamori Labs, LLC · Voice & Language · Productivity · AI AssistantFoley is a voice to-do list that turns a spoken sentence into a filed task, complete with project, date and time, then helps you plan the day around what is already there.
HireKey
HireKey · Voice & Language · Productivity · LeaningHireKey is an AI job search copilot built by a former tech recruiter. It checks role fit, tailors resumes, runs interview practice and tracks every application.
Fluence
Yay Software (Yannick Huchard) · Writing · Voice & Language · ProductivityFluence is an Android app that reads PDFs, EPUBs, Word files and web articles aloud, follows each sentence as it speaks, and answers questions about the text on the phone itself.