Speech
Speech generation, recognition, voice interfaces, and real-time audio systems.
Applied AI laboratory
Praha Lab
Beyond Artificial. Toward Intelligence.
We develop AI models and the systems required to evaluate, deploy, and use them across speech, language, agents, and multimodal intelligence.
Research domains
One lab. Multiple model systems.
Praha Lab works across the layers between model research and deployed intelligence.
Language
Language models, reasoning systems, evaluation, and domain-specific intelligence.
Agents
Tool-using systems that can plan, act, remember, and operate across workflows.
Multimodal
Models and products that work across text, audio, vision, and interaction.
Release 001 / Speech
RimaTTS V1
A multilingual Indian text-to-speech model with voice cloning, built for natural speech across eight languages.
Praha Lab / Model 001RimaTTS V1
RimaTTS generates speech and clones a speaker's voice from a short reference recording. It is the first public model from Praha Lab.
- Languages
- 8 Indian languages
- Capability
- Voice cloning
- Version
- V1
- Access
- Samples pending
- Hindi
- Bengali
- Marathi
- Gujarati
- Tamil
- Telugu
- Malayalam
- Kannada
The lab
Models are one part of the system.
Our work combines research, infrastructure, and product engineering. Each layer informs the next.
Research
Model architecture, training, evaluation, and the failure modes that determine whether a system is useful.
Infrastructure
Inference, latency, deployment, and the operational layer required to run models beyond a notebook.
Products
Focused interfaces and APIs that make model behavior observable, testable, and accessible.
The lab
Research. Systems. Intelligence.
For model access, evaluation, research, or technical collaboration.
Contact the lab