IN
Inworld AI
Low-cost realtime voice infrastructure for AI applications
Voice & Speech AI
usage based
Verified Sep 2026Overview
Inworld has repositioned away from its earlier AI character/NPC focus toward realtime inference infrastructure.
Key features
- Realtime TTS with sub-100ms latency, voice cloning and 200+ languages
- Realtime STT with voice profiling for emotion, age, accent and pitch
- Realtime API over WebSocket with function calling
- Realtime Router
Best for
Developers optimizing cost and latency in voice pipelines
Specifications
| Pricing model | usage based |
|---|---|
| Pricing detail | Realtime TTS quoted at $12.50 per million characters; Realtime STT at $0.10/hour; Realtime Router offers routing across 220+ LLMs at zero markup. |
| Deployment | cloud |
| API | Yes |
| Free tier | No |
| Integrations | — |
| Company | — |
| HQ | — |
| Notable clients | — |
Verification
Source: https://www.inworld.ai · Last verified Sep 17, 2026
Report an inaccuracyAlternatives
Text-to-speech, voice cloning and voice agents
freemium
API
Free tier
Source: elevenlabs.io · Last verified Sep 17, 2026
Speech generation and recognition APIs
API
Free tier
Source: cartesia.ai · Last verified Oct 19, 2026
Speech recognition and synthesis APIs
API
Source: deepgram.com · Last verified Oct 19, 2026
Text-to-speech with streaming and voice controls
API
Free tier
Source: wellsaid.io · Last verified Oct 19, 2026