Overview
Eclatira is a real-time conversational AI platform for building multimodal agents that combine native voice-to-voice interaction with live camera and screen input in a single bidirectional session. Where most voice AI asks a customer to email a photo and wait, Eclatira sees the passport, the label, or the broken part mid-call, OCRs it, and acts — booking a room, starting an RMA, or verifying ID without a human handoff. The product launched through Product Hunt in late 2026 and targets front-office and customer-facing operations: hotels, clinics, retail and support lines where a call, a walk-in, and a held-up object are part of the same transaction. Builders can describe an agent’s job in plain language or use a versioned REST API, and connect to their own APIs, MCPs, and 3,000+ third-party apps. Voice and video run at up to 30 frames per second within a sub-800ms latency budget, keeping visual understanding in sync with the conversation.
Key Features
- Voice and vision in one stream — audio and live video (webcam or screen share) flow through a single pipeline, so the agent discusses what it sees as you talk instead of requiring a separate upload-and-wait step.
- Live OCR and object recognition — it reads printed pages, on-screen text, and product labels in view and acts on them, which is the feature that separates it from blind voice bots.
- Action and integrations — connect to booking engines, CRMs, and any API; Twilio and Telnyx telephony are built in, and calls end with the work already done.
- Sub-800ms, 100+ languages — full-duplex voice with barge-in across 100+ languages and 300+ regional accents keeps interactions natural and global.
- No-code builder plus REST API — describe the job in plain English or programmatically create agents, upload docs, define tools, and start calls.
- Privacy and deployment options — encryption in transit and at rest, workspace isolation, configurable retention, and an enterprise tier with isolated infrastructure and SLA.
Pricing
| Plan | Price | For |
|---|---|---|
| Free | $0 | 1,350 credits/mo (~30 voice min), 2 concurrent calls, unlimited agents |
| Solo | ~$35–49/mo | 7,000–12,000 credits, 8 concurrent calls, telephony, email support |
| Growth | ~$990/mo | 50 concurrent calls, 100MB KB, priority support |
| Enterprise | Custom | Unlimited credits, dedicated infra, SLA, compliance |
Credits are the single usage currency: ~45 per voice minute, ~25 per text message. Exact Solo/Starter/Scale figures vary across public listings — confirm on eclatira.com/pricing.
Comparison
vs. HeyGen: HeyGen is an avatar video generator for scripted, one-to-many marketing and training clips. Eclatira is the opposite end — live, two-way conversational agents that react to what a customer shows the camera, built for support rather than content production.
vs. Synthflow: Synthflow is a no-code voice-agent platform for phone and web. Eclatira adds a real visual layer — camera and screen — so it can handle tasks where seeing the object or screen is the whole point.
vs. Dial: Dial gives an AI agent a real phone number and SMS across 200+ countries via a single API. Eclatira is a broader multimodal agent platform (voice, video, screen, chat) rather than a telephony-first primitive.