New accounts get 1,000 free credits in the studio. No card required.

Models

An ever-expanding catalogue of models

View all models
Deepgramagent
Deepgram Voice Agent
from $0.050 / min
Deepgramspeech-to-text
Deepgram Nova-3
$0.0043 / min
Deepgramtext-to-speech
Deepgram Aura-2
$30 / 1M chars
Pascalspeech-to-text
Pascal Transcribe
$0.18 / hr
Tencentvoice-cloning
AuK Flash Clone
$0.0042 / min
Tencentvoice-cloning
AuK Base Clone
$0.0225 / min
OpenAIspeech-to-text
Whisper Large v3 Turbo
$0.0015 / min
NVIDIAspeech-to-text
Parakeet TDT 0.6B v3
$0.00133 / min
NVIDIArealtime ASR
Nemotron 3.5 Streaming
$0.00333 / min
Sarvamspeech-to-text
Sarvam Saaras
$0.005 / min
Sarvamtext-to-speech
Sarvam Bulbul
$30 / 1M chars
Piperrealtime TTS
Piper TTS Realtime
$0.00333 / min
Matcharealtime TTS
Matcha TTS Realtime
$0.00333 / min
Resembletext-to-speech
Chatterbox Turbo
$0.0075 / min
Qwentext-to-speech
Qwen3-TTS 1.7B
$0.020 / min
hexgradtext-to-speech
Kokoro
$0.001 / min
k2-fsatext-to-speech
OmniVoice
$0.009 / min
Metastem-separation
HTDemucs
$0.00117 / min
Deepgramagent
Deepgram Voice Agent
from $0.050 / min
Deepgramspeech-to-text
Deepgram Nova-3
$0.0043 / min
Deepgramtext-to-speech
Deepgram Aura-2
$30 / 1M chars
Pascalspeech-to-text
Pascal Transcribe
$0.18 / hr
Tencentvoice-cloning
AuK Flash Clone
$0.0042 / min
Tencentvoice-cloning
AuK Base Clone
$0.0225 / min
OpenAIspeech-to-text
Whisper Large v3 Turbo
$0.0015 / min
NVIDIAspeech-to-text
Parakeet TDT 0.6B v3
$0.00133 / min
NVIDIArealtime ASR
Nemotron 3.5 Streaming
$0.00333 / min
Sarvamspeech-to-text
Sarvam Saaras
$0.005 / min
Sarvamtext-to-speech
Sarvam Bulbul
$30 / 1M chars
Piperrealtime TTS
Piper TTS Realtime
$0.00333 / min
Matcharealtime TTS
Matcha TTS Realtime
$0.00333 / min
Resembletext-to-speech
Chatterbox Turbo
$0.0075 / min
Qwentext-to-speech
Qwen3-TTS 1.7B
$0.020 / min
hexgradtext-to-speech
Kokoro
$0.001 / min
k2-fsatext-to-speech
OmniVoice
$0.009 / min
Metastem-separation
HTDemucs
$0.00117 / min

API and Studio, for what you need

API

Pay as you go.

Top up $10 or more at once and get an additional $10 in API credit within 30 days of signup.

API Console

Studio

Web interface for models. Subscribe for monthly studio credits.

Upload, run, download.

Unified API. Change the model name to switch models.

curl https://api.pascalai.co/v1/speech-to-text \
  -H "Authorization: Bearer $PASCAL_API_KEY" \
  -F model=whisper-large-v3-turbo \
  -F file=@input.wav

How it works

One unified setup. The same model names, parameters, and outputs across Studio and the API.

1
Pick a model

Choose from 40+ models across Voice generation, Transcribe, Localisation, Separation & enhancement, and Agents. Same IDs in the studio and on the API.

01 · Catalog Selector
Studio & API
Voice generationTranscribeLocalisationSeparationAgents
TTS$0.001 / min
Natural text-to-speech voice generation
voice: natural-alloyspeed: 1.0format: 24kHzstream: true
TTS · Expressive$0.0012 / min
Conversational multi-speaker speech synthesis
voice: studio-clarityspeed: 1.0format: 24kHzmultilingual
✓ Direct API & Studio parityMore models ↓
1

Pick a model

Choose from 40+ models across Voice generation, Transcribe, Localisation, Separation & enhancement, and Agents. Same IDs in the studio and on the API.

2

Run it

Send a standard JSON payload via API or click synthesize in the browser Studio. Zero complex setup, GPU cold starts, or infra management.

3

Get the result

Stream real-time audio, receive word-accurate timestamps, or download separated audio stems in lossless formats.

Studio plans

Creator

For individuals producing regularly.

$10first month

Then $20/month

20,000 studio credits / month

  • All available models
  • Standard queue
  • Commercial usage rights

Max

For high-volume studio production.

$100/month

100,000 studio credits / month

  • Everything in Pro
  • Higher concurrency
  • Highest monthly credit allowance

Business

Talk to us about your requirements

Custom
  • Custom SLA
  • On-premise deployment
  • Dedicated support
  • Volume and custom contracts
Contact us

Every audio workload ripples across the map. Use the API or Studio to put your city on the map.

1,226hours of audio routed
14cities on the map
Global Audio Activity MapBengaluru, IndiaVoice generationDelhi, IndiaSpeech to textMumbai, IndiaVoice agentSan Francisco, United StatesSpeech to textNew York, United StatesVoice generationChicago, United StatesVoice agentAustin, United StatesAudio enhancementLondon, United KingdomVoice generationParis, FranceSpeech to textFrankfurt, GermanyVoice agentTokyo, JapanSpeech to textSingaporeVoice generationSydney, AustraliaAudio enhancementSão Paulo, BrazilSpeech to text
Use the API or Studio to put your city on the map.
Transcribe 532 hrs Voice generation 535 hrs Voice agents 9 hrs Separation & enhancement 122 hrs Localisation 28 hrsleads
Listening for live audio routes... Try generating or uploading audio above.

FAQ

What does Pascal do?

One API across speech categories, with a working setup for each model. Changing model is a name change. When a new model ships, you stay current without doing the setup again.

Who is it for?

People who want the best model available and don't want to rebuild every time a new better model comes out.

Are Studio and the API the same thing?

No. They are different. A subscription is for Studio — monthly credits in the browser. The API is pay-as-you-go.

Do I need a studio plan to use the API?

No. The API is pay-as-you-go. It starts at zero. Add any amount from $5. Studio plans are only for monthly credits in the browser.

What are credits?

Usage is prepaid software access. One unit equals $0.001. Studio plans include monthly studio usage. API purchases add separate API usage. Keys cannot spend studio usage. Model list prices are shown in dollars. Usage is not transferable and has no cash value.

What is included in a studio plan?

A monthly credit allowance for studio tools. Same models as the API.

Do unused plan credits roll over?

No. Studio plan usage resets each billing cycle. Unused usage does not carry forward.

Do new accounts get free credits?

Yes. New accounts start with 1,000 credits in the studio. The API starts at zero. Within 30 days of signup, one API top-up of $10 or more adds $10 of credit.

How is usage billed?

A successful job includes a $0.05 fee plus the price on the card. The card unit is the meter: per minute of audio, per audio hour, or per million characters. Dubbing is per second of source audio for each target language. API credit is added from $5. The platform fee is 5.5% or $0.80, whichever is larger.

Can I switch models without starting over?

Yes. Pick another model from the catalog and run the same job again.

Is my data used to train models?

No. None of your uploads, scripts, or generated audio are used to train any model.

Ready to try a model?

Open the studio and run your first job in the browser.