A text-to-speech API built for content automation and app services to conversational agents. Integrate AI voices with nuanced emotional expression wherever you need them.
*No payment info required
curl --request POST \
--url https://api.typecast.ai/v1/text-to-speech \
--header 'Content-Type: application/json' \
--header 'X-API-KEY: <api-key>' \
--output output.wav \
--data @- <<EOF
{
"voice_id": "tc_60e5426de8b95f1d3000d7b5", Smart Emotion
Smart Emotion uses AI to automatically understand context then apply the most fitting emotion. Best for services that need instant responses, such as conversational agents.
Oh my gosh, we actually did it and it finally worked out! This is the best news ever, and I feel like I could jump up and cheer for an hour!
from typecast import Typecast
from typecast.models import TTSRequest, SmartPrompt
client = Typecast(api_key="YOUR_API_KEY")
response = client.text_to_speech(TTSRequest(
text="Oh my gosh, we actually did it and it finally worked out!",
model="ssfm-v30",
prompt=SmartPrompt(emotion_type="smart")
))
with open('typecast.wav', 'wb') as f:
f.write(response.audio_data)Key features
Integrate natural-sounding TTS through one simple API, with everything you need to build, launch, and scale.
AI understands a text's surrounding context and automatically adds the right emotional tone. You can also fine-tune it manually with 7 emotion presets.
"prompt": { "emotion_type": "smart", "previous_text": "I feel like I'm walking on air", "next_text": "I am literally bursting with happiness"
From energetic short-form video to calm narration and kids' content, choose from a wide range of unique voices across different ages, tones, and personalities. (View voices)
Integrate real-time services such as conversational agents and chatbots with a Streaming TTS API that responds in 200ms.
Natural English comes standard, with support for global services in 35+ languages. (See supported languages)
Create an AI voice clone from an audio sample as short as 5 seconds. Streamline production with your own AI voices.
Supports a wide range of integrations, from Python and JavaScript SDKs and CLI to no-code automation tools like Zapier and n8n.
From TikTok and YouTube Shorts voiceovers to calm announcer-style narration, we have 500+ voices ready for you.
Use cases
From social media content to conversational AI, integrate natural voices with our text-to-speech API
Content automation
Narration for YouTube, documentaries, ads, and more
Automatic subtitles
Generate word/character-level timestamps
Audiobooks and podcasts
Expressive narration with realistic emotion
Conversational agents
Naturally voiced chatbots and AI assistants powered by Smart Emotion
AI tutors
Distinctive educational voices tailored to each learner
AICC
Reliable 24/7 responses with human-like natural voices
Why Typecast
Open-source TTS can help you add voices quickly, but it takes more to create speech that feels natural, expressive, and engaging.
| Category | Typecast (SSFM 3.0) | Open-source TTS (xtts v-2) | Company E (Multilingual v2) |
|---|---|---|---|
| Unit cost | Plans include monthly credit allowances and are offered at around 20% lower unit cost than global TTS APIs. | The model is free, but GPU infrastructure and ML engineering costs are covered by the user. | Usage-based pricing, with relatively higher unit costs for the Multilingual model. |
| High-quality voices | Instantly use 500+ verified voices designed by professional voice actors in the industry. | Voiceover variety is limited. | Many voices are user-generated, with only a few that are curated and high-quality. |
| Natural emotional expression | Smart Emotion automatically adjusts emotion and intonation to match the content. Available for real-time services. | Emotion is neutral, and conversational use cases require additional work. | Emotion can be controlled manually by entering audio tags. |
| Real-time language support | Supports 35+ languages and provides native-level naturalness in 6 languages, including English. | Supports 17 languages, with significant variation in naturalness by language. | Supports 32 languages, mainly English-focused, with limited naturalness in other languages. |
| Streaming latency | Around 200ms TTFB for text-to-speech API streaming, stable enough for real-time conversations and live services. | Varies depending on your GPU environment and model; no official figure is available. | Varies by model. Multilingual models have around 250-300ms TTFB and are stable enough for real-time conversations and live services. |
| Voice cloning | Instant Cloning is available with voice samples ranging from 5 to 150 seconds. | Zero-shot cloning is possible with 6 seconds of audio, with an average reproduction quality. | Offers instant and professional options, so you can choose between faster production or higher quality depending on your use case. |
Model
The naturalness of Typecast's SSFM model is the result of AI voice research and proprietary model development dating back to 2017.
Introduced deep learning-based speech synthesis technology
Improved phoneme-based pronunciation accuracy and naturalness
Introduced a large-scale data framework
Expanded multilingual support and added attribute control
Context-based emotion control, trained on millions of hours
Pricing
Get started in minutes with transparent usage-based pricing, or work with our team to build a custom plan for large-scale deployments.
Free
Lite
Plus
Enterprise
"output": {
"volume": 100,
"audio_pitch": 0,
"audio_tempo": 1,
"audio_format": "wav"
},
"seed": 42
} Tips, tricks and tools for creators, developers, and storytellers.
Typecast Pricing Update & Scheduled Maintenance on July 31
Hyelee Seo | Jul 24, 2026
How to Use AI at Work: A Complete Guide
Joe Crosby | Jun 23, 2026
AI Business Intelligence: How to Use AI for Data-Driven Decisions
Joe Crosby | Jun 17, 2026
How to Build a Second Brain System
Joe Crosby | Jun 15, 2026
Find the Perfect Voice Faster With This AI Voice App
Joe Crosby | Jun 10, 2026
Voiceover for PowerPoint: 7 Best Methods That Work
Hyelee Seo | Jun 9, 2026
Start integrating context-aware emotion, expressive voices, and low-latency speech streaming into your online content, conversational AI, and app services.
*No payment info required
