A research lab and inference provider for realtime voice AI at consumer scale. Inworld offers first-party TTS and STT models (Realtime TTS-2 and TTS-2 Flash), model-agnostic routing across 100+ LLMs with zero markup, speech-to-speech orchestration, and built-in observability—all in one vertically integrated stack. Realtime TTS-2 ranks #1 on Artificial Analysis with sub-100ms time-to-first-byte, while TTS-2 Flash pushes latency under 25ms for high-volume applications. The platform serves three verticals: consumer apps (companions, social, language learning—Status by Wishroll hit 1M users in 19 days with 95% AI cost reduction), enterprise voice agents (support, sales, recruiting), and interactive media (games and entertainment, with partners including Xbox, NVIDIA, Ubisoft, and Niantic). Pricing is credit-based with a free On-Demand tier for prototyping; LLM routing passes through provider costs with no markup.
Pros & Cons:
Pros: First-party TTS-2 model is top-ranked on Artificial Analysis with sub-100ms TTFB, and TTS-2 Flash offers sub-25ms for latency-critical apps ; model-agnostic LLM routing across 100+ models with zero markup on provider costs ; vertically integrated stack (TTS + STT + speech-to-speech + routing + observability) reduces glue-code overhead; proven at scale with consumer apps and game studios (Xbox, Ubisoft, Niantic) ; free On-Demand tier for prototyping with up to 70 minutes of TTS included .
Cons: Primarily a developer/infrastructure product—no end-user app or no-code interface; pricing scales with usage and can become significant at volume (Growth at $1,500/mo for large deployments) ; consumer app builders still need to manage their own product layer, memory, and UX; the platform\'s origin in gaming/characters may make it less immediately obvious for non-entertainment use cases; third-party independent reviews on G2/Capterra are not surfaced in search results.
**Who is it for?**:Developers and product teams building realtime voice-first consumer apps (companions, social, language learning); enterprises deploying voice agents for support, sales, or internal workflows; game studios and interactive media companies that need scalable, expressive character voices with lip-sync support; AI agent builders who want a single provider for TTS, STT, and LLM routing rather than stitching together multiple vendors.