Model Terminal

Maya1

Maya1 is an open-source 3B-parameter text-to-speech model from Maya Research that generates expressive, emotionally controllable speech from text and natural-language voice prompts. It outputs 24 kHz mono audio using SNAC codec tokens and is designed for real-time, single-GPU deployment. Source: supabase.

Value score
Not scored
Context
tokens
Max output
tokens
Price

Capability radar

Identity

Developer
Maya Research
Openness
Modalities
Release
Knowledge cutoff
Deprecation
API docs

Benchmarks

No benchmark scores yet.

Compare nearby

Pricing

Input / 1M
Output / 1M
Speed