Loquere.
Own-voice evidence for every student. After a unit, each student retells it to a native-voiced character (a Mexican abuela, a bilingual Chicago college student), and you get their audio plus a transcript you can trust. Runs on our hardware, no student accounts. Optional: the base tier never depends on it.
Voices from across the Spanish-speaking world.
The conversation partner cast for Latin is being recorded now. Drop your email and we will let you know when the first personas ship.
Built specifically for K-12 World Language. Not for adult app users.
ElevenLabs voice cloning calibrated per persona. A Mexican abuela does not sound like a Madrid teen. A Cuban-American college student does not sound like a Buenos Aires professor. Dialect is not a setting; it is the persona.
Live2D avatars with Wawa-Lipsync. Real-time mouth animation matched to phonemes. Not a static portrait, not an uncanny 3D head. Editorial illustration that moves with the voice.
AS-units analysis for fluency, embedding classifiers for proficiency level, LanguageTool plus a fine-tuned model for grammar feedback. Specific, actionable, never drive-by corrective.
Students join with a class code, first name only. No email collection from minors. Conversation transcripts live in the teacher dashboard, not in a vendor data lake. FERPA-safe by design.
Be first in line for the classroom pilot.
We will email you when the first cohort opens. Pilot teachers get founding-member Pro pricing locked in for life, plus a say in which personas ship first.