Ablero

Real-time voice AI for the Spanish-speaking world.

Ablero is a voice AI lab building speech models that hold a conversation the way people do: listening while they speak. We're starting with the hardest audio there is, the phone call.

What we're building

We're building open speech models for real-time conversation in Spanish.

First is Ablero ASR, streaming speech recognition for phone calls. It's built for 8 kHz G.711 audio and for the details a call turns on: names, ID numbers, dates.

Next is Ablero Duplex, a full-duplex model that listens while it speaks, handles interruptions and hands tasks to your agents mid-conversation.

Then come voice agents, built end to end on our own models.

What we hold to

• Real time is the product. Latency is what the caller feels, so we design for it first.
• Spanish first. Built for Spanish across Latin America and Spain, not translated from English. Catalan, Basque and Galician are on the roadmap.
• Open weights, hosted inference. We release our models openly here. Our API runs them in production for you.
• Built in Europe. Hosted in the EU or inside your network. Call audio stays under your control.

Work with us

We're onboarding our first pilot partners: call centres, BPOs and voice-agent teams across Latin America and Spain.

ablero.dev · igor@ablero.dev