Skip to Content
Models

Models

Revolab offers TTS models (Nada) and STT models (Aisyah). The tables below list the models this environment serves right now. They are read live from the platform, so they never go stale. List them programmatically with GET /v1/models (OpenAI list shape — client.models.list() works).

All prices are in Malaysian Ringgit (RM) per audio minute processed, and are read live from the platform’s current pricing.

TTS models (Nada)

Nada models synthesize speech from text. Pass the model ID as the model field in POST /v1/tts. If you omit model, the platform uses its default TTS model, which is always one listed here.

Model IDNameBest forPriceUnit

Loading the models this environment serves…

STT models (Aisyah)

Aisyah models transcribe audio to text. Pass the model ID as the model form field in POST /v1/stt. If you omit model, the platform uses its default STT model, which is always one listed here.

Model IDNameBest forPriceUnit

Loading the models this environment serves…

Billing

Billing is based on audio duration — the length of the generated audio (TTS) or of the uploaded audio file (STT). You pay for the exact duration, with no rounding up to the next second or minute: 4.6 seconds of audio is billed as 4.6 seconds.

Free allowances

The figures below are read live from the platform. A model’s price in the tables above shows which kind of allowance applies: Free or an RM price.

Paid models. Every new account gets … minutes of STT and … minutes of TTS to spend on the paid models. This allowance is granted once, does not expire, and does not reset monthly.

Each pool is shared across all paid models of that type, and a minute costs the same against the pool whichever paid model you call. The free models never draw it down.

Once a pool is used up, that model type bills against your MYR balance at the rates above; if the balance is empty, requests return 402 insufficient_balance.

Choosing a model

  • Interactive / real-time apps (chatbots, IVR): choose the fastest model listed — a Flash or Flash Lite model where one is offered.
  • Premium content (audiobooks, voiceovers, legal transcription): choose a Pro model — the highest quality.
  • Each model’s “Best for” column above summarizes its trade-off.
  • Rate Limits & Quotas — per-key RPM limits and how to handle 429 responses.
VoicesErrors