What Inworld AI does

Inworld AI (legal entity Theai, Inc.) is a Mountain View-based voice AI infrastructure company founded in 2021 that builds realtime models and tooling for building AI agents with natural conversational voice. The company's product portfolio centers on four core APIs: Realtime TTS, ranked #1 on the Artificial Analysis Speech Arena and shipping with sub-130ms latency, voice cloning from 15 seconds of reference audio, and cross-lingual support across 100+ languages; Realtime STT with proprietary voice profiling that extracts emotion, accent, age, pitch, and vocal style alongside transcription; Realtime API providing end-to-end speech-to-speech over WebSocket with tool calling and dynamic context management; and Realtime Router, an intelligent LLM router that spans 220+ models from OpenAI, Anthropic, Google, xAI, Mistral, and Groq at zero markup. Beyond APIs, Inworld Runtime (launched August 2025) extends the company into a broader AI runtime layer for consumer applications.

The technology stack rests on a proprietary closed-loop voice architecture (TTS-2) that ingests prior-turn audio to modulate tone, pacing, and emotional state—fundamentally different from text-conditioned baselines—and is offered alongside full-stack integrations with LiveKit, Vapi, Pipecat, and Mastra. Revenue is generated through usage-based API pricing (TTS at $0.10 per million characters, STT at $0.10 per hour), tiered subscriptions including a free tier and enterprise Max plans with dedicated GPU infrastructure at $5/GPU-hour, and direct enterprise sales to gaming and platform customers. The go-to-market combines an API-first self-serve platform (platform.inworld.ai) with field sales into AAA gaming and major platforms including Xbox/Microsoft, Ubisoft, NVIDIA, Niantic, and Streamlabs, complemented by a developer relations program across GitHub, Discord, X, LinkedIn, and GDC.

The customer base is bifurcated: enterprise anchors (Xbox, Ubisoft, NVIDIA, Niantic, Streamlabs) provide strategic validation and gaming-industry distribution, while a long tail of consumer AI startups (Talkpal AI at 5M users, OtherHalf at 1M users in 19 days, Death by AI at 20M players, Wishroll at 500K+ DAU) drives usage-based volume. Inworld is SOC 2 Type II certified, HIPAA compliant, GDPR and CCPA compliant, and references alignment with the EU AI Act, supporting expansion into regulated verticals such as healthcare, education, and contact centers. Leadership is led by co-founder and CEO Kylan Gibbs; cumulative funding exceeds $117M across seed, Series A, and Series A2 rounds from Kleiner Perkins, CRV, Meta, BITKRAFT Ventures, Intel Capital, Lightspeed, Microsoft M12, Founders Fund, Stanford, and others.

Name

Inworld AI

Legal name

Theai, Inc. dba Inworld AI

Website

https://inworld.ai

Company type

Private

Founded year

2021

Operating status

Operating

Headcount range

101–250 employees

Short description

Inworld AI builds realtime voice AI infrastructure—text-to-speech, speech-to-text with voice profiling, speech-to-speech APIs, and an LLM router across 220+ models—used by consumer AI startups (Talkpal, OtherHalf, Death by AI) and enterprise gaming and platform customers (Xbox, Ubisoft, NVIDIA, Streamlabs).

Ownership category

akta.pro rank

Product category

Voice AI Infrastructure

NAICS

Software Publishers (51321), Computer Systems Design Services (541512)

SIC

Services-Prepackaged Software (7372)

akta.pro primary industry

Conversational AI Platforms (chat/voice bots, orchestration) (HDAAAFAC)

akta.pro secondary industries

Voice Assistants & Voice UI (VUI) (HDAAAFAD), Model Serving, Inference & Deployment Platforms (APIs, Edge/On-Prem) (HDAEANAC), Enterprise Foundation Model Integration & APIs (Connectors, Governance, Deployment) (HDAAACAO)

Keywords

• Realtime voice AI
• Text-to-speech API
• Speech-to-text API
• LLM routing platform
• AI character infrastructure

Headquarters

HQ city

Mountain View

HQ country

United States

HQ region

North America

Offices6 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales

Revenue model

1. API Usage-Based Revenue: Usage-based pricing for TTS (per million characters), STT (per hour), and LLM inference (percentage of provider rates). Scales with customer usage. Customers can start free and scale based on usage.
2. Subscription Plans: Tiered subscription plans including Growth and Max plans with different rate limits, features, and pricing. Includes dedicated GPU options at enterprise scale.

Pricing tiers

Go-to-market motion2 records

Distribution channels4 records

Marketing channels10 records

Core offering

Inworld AI provides a realtime voice AI infrastructure platform comprising text-to-speech (TTS), speech-to-text (STT), end-to-end Realtime API, and an LLM routing layer (Realtime Router) that connects developers to 220+ foundation models with zero markup. The platform is targeted at building emotionally aware, low-latency voice AI applications for consumer AI, gaming NPCs, language learning, healthcare companions, and enterprise customer service. Inworld also offers the Inworld Runtime, an AI runtime layer for consumer applications that enables AI characters to learn and adapt from user interactions.

Product overview

Inworld AI is a research lab building realtime voice AI infrastructure, positioned as the #1 ranked realtime voice AI platform. The product portfolio centers on four core API products: Realtime TTS (text-to-speech), Realtime STT (speech-to-text with voice profiling), Realtime API (end-to-end speech-to-speech), and Realtime Router (LLM routing across 220+ models). These integrate into a unified pipeline for building voice AI applications. Supporting infrastructure includes Inworld Runtime (AI runtime for consumer applications) and Inworld CLI (developer tools). The platform targets consumer AI applications including companions, agentic workforce, learning & education, health & wellness, and interactive media. Pricing emphasizes cost reduction with over 50% savings versus competitors, positioning as consumer AI infrastructure optimized for scale.

Differentiator

Problem solved

Functional benefit

Products and services

• Realtime TTS Text-to-speech API ranked #1 on Artificial Analysis Speech Arena. Provides sub-130ms first-chunk latency, voice cloning, cross-lingual support across 100+ languages, and advanced voice direction via plain-language prompts. Powers natural, realtime conversational voice output for developers and enterprises.
• Realtime STT

Quantifiable outcome

• Reached 1M users in 19 days (OtherHalf)
• +6 more outcomes

Named customers18 records

Segments6 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration14 records

AI capability8 records

Feature6 records

Partnerships

Ten partnerships are on record, tiered flagship, major and core.

• Xbox / MicrosoftflagshipStrategic or Co-development Partner · 25 June 2025Partnership to build AI tools for game development announced June 2025. Deep integration with Microsoft gaming ecosystem.
• StreamlabsmajorTechnology or Integration · 6 January 2025Collaboration with NVIDIA and Inworld AI announced at CES 2025 for AI-powered streaming tools. Integration of Inworld voice AI into Streamlabs platform.
• UbisoftflagshipStrategic or Co-development Partner · 1 March 2024Partnership with NVIDIA and Inworld AI to develop NEO NPCs with dynamic conversation capabilities. Major AAA game studio partnership for AI-powered NPCs in Far Cry and other titles.
• The Walt Disney CompanymajorStrategic or Co-development Partner · 14 July 2022Selected for Disney Accelerator 2022 program focusing on immersive experiences using AR and AI. Received investment, mentorship, and access to Disney's Los Angeles campus for development.
• NVIDIAflagshipStrategic or Co-development PartnerStrategic partnership for AI NPCs using NVIDIA ACE (Avatar Cloud Engine). Inworld AI is a key partner in NVIDIA's AI gaming ecosystem for dynamic NPC technology.
• LiveKitmajorTechnology or IntegrationTechnology integration combining Inworld TTS-2 with LiveKit agents for emotionally expressive voice synthesis. LiveKit provides realtime communications infrastructure.
• VapimajorTechnology or IntegrationVoice agent platform integration with Inworld for building expressive, real-time voice agents. Vapi provides voice AI framework infrastructure.
• PipecatmajorTechnology or IntegrationPartnership raising the bar for realtime voice AI through Pipecat's voice agent framework integration.
• ModularcoreTechnology or IntegrationCollaboration to make state-of-the-art speech synthesis scalable. Modular provides AI inference optimization for Inworld's voice models.
• Mistral AImajorTechnology or IntegrationCustom Mistral AI models integrated with Inworld Runtime for enhanced engagement and performance improvements in AI character applications.

Scale indicators10 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• ElevenLabs: Leading voice AI / TTS platform focused on high-quality speech synthesis, voice cloning, and dubbing. Direct competitor in TTS pricing and quality (Inworld explicitly benchmarks 50%+ cost reduction vs ElevenLabs).
• Resemble AI: Voice AI platform specializing in voice cloning, TTS, and real-time speech-to-speech. Direct competitor in voice cloning, emotion-aware synthesis, and enterprise voice AI deployments.
• Vapi: Voice agent platform for building conversational AI agents. While also a technology partner, Vapi competes for the same developer mindshare in building realtime voice agent applications.
• Deepgram: Speech-to-text / voice AI platform offering real-time transcription and voice agent APIs. Direct competitor in STT pricing (Inworld benchmarks 50x cost reduction vs Deepgram) and increasingly in full voice agent stacks.
• PlayHT: AI voice generator and TTS platform with voice cloning and conversational AI offerings. Direct competitor in TTS quality, voice cloning, and developer API pricing.

Broad incumbents

• OpenAI: Realtime API and TTS/STT offerings bundled with GPT-4o and broader foundation models. Competes as upstream provider (via Inworld Router) and as potential competitor in full voice agent stacks.
• Google Cloud Speech: Cloud-based TTS, STT, and conversational AI offerings bundled with Gemini models. Hyperscaler competitor with significant R&D resources and existing enterprise distribution.
• Microsoft Azure Speech: Cloud speech services (TTS, STT, speaker recognition) bundled with broader Azure enterprise stack. Competes via enterprise procurement and bundled contracts; M12 (Microsoft's VC arm) is also an Inworld investor.

Emerging players

• Hume AI: Voice AI platform with emotionally intelligent speech-to-speech (Empathic Voice Interface). Direct overlap in emotion-aware voice AI positioning and consumer AI applications.
• Cartesia: Real-time voice AI startup focused on ultra-low-latency speech models (Sonic). Emerging direct competitor in realtime TTS and conversational voice agents with strong technical positioning.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights7 records

Customer concentration

• Xx.com/inworld_ai
• GitHubgithub.com/inworld-ai
• LinkedInlinkedin.com/company/inworld-ai

Compliance5 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles3 records

Funding overview

Funding rounds6 records

Investors25 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Inworld AI