Inworld AI Blog the company blog of Inworld: latest posts and feed
Inworld · Company blog · 1 a month · last post Oct 14, 2026 · Active
Posts
26 on file since Jun 1, 2025AI and agentsOct 14 10d10 days agoIntroducing Inworld InferenceWe rebuilt how open models are served for consumer apps, from the attention kernels to the cache router. Today it is open to every developer.Developer toolsSep 30 3w3 weeks agoWe have acquired Ultravox to Advance Realtime Voice AIInworld has acquired Ultravox, the platform developers use to build real-time voice agents. Built-in Inworld voices on Ultravox now run on Realtime TTS-2, with no code changes required.Product and companyAug 31 7w7 weeks agoRealtime TTS-2: A new frontier voice model that feels as human as it soundsA new frontier voice model that feels as human as it sounds. Hears the conversation, takes voice direction in plain English, and holds one voice across over 100 languages.Product and companyAug 28 8w8 weeks agoIntroducing the Inworld TTS Open Evaluation ToolkitAn open source framework for reproducible TTS evaluation. Local ASR, explicit normalization, per-sample measurements, and offline reports you can trace back to the audio.AI and agentsJun 15 4mo4 months agoBuilding conversational voice agents with Mastra + Inworld Realtime API@mastra/voice-inworld-realtime collapses STT, LLM, and expressive realtime TTS into one WebSocket, with semantic VAD, barge-in, and tool calling out of the box.AI and agentsJun 10 4mo4 months agoCost is the wall in front of consumer AI. We are taking it downConsumer AI was priced for the enterprise, and that arithmetic is where good products die. We are cutting the cost of the whole stack in half or more for most developers, across text-to-speech, speech-to-text, the LLM, and compute, and...AI and agentsMay 14 5mo5 months agoRealtime voice agents can now see, listen, and engageInworld TTS-2 + Stream Vision Agents build AI voice agents that read emotions, respond in real time, sub-200ms latency, 100+ languages, full context.Reliability and performanceMar 13 7mo7 months agoBeyond Quality: Emotionality and ExpressivenessStandard TTS benchmarks miss emotional depth. Inworld's framework measures arousal and prosodic variation, the metrics that actually predict user engagement.Reliability and performanceJan 21 9mo9 months agoInworld TTS-1.5: Upgrading the #1 Ranked TTS Model with Production-Grade Latency, Expression and StabilityTTS-1.5 Max hits P90 under 250ms. Mini under 130ms. 40% fewer word errors, 30% more expressive, 15 languages including Hindi. Built for production scale.ResearchJan 13 9mo9 months agoTalkpal AI scales to 5 million language learners with Realtime TTSHow Talkpal cut TTS costs by 40% while improving retention and free-to-paid conversionResearchNov 6, 2025 11mo11 months agoIntroducing timestamp alignment, WebSockets and more for Realtime TTSNew Realtime TTS features: custom pronunciation APIs, word and character timestamps for lipsync, WebSocket streaming, Russian language support, and quality improvements.AI and agentsNov 4, 2025 11mo11 months agoThe 3 Engineering Challenges of Realtime Conversational AIBuilt from our learnings in latency and iteration speed, Inworld Runtime is the C++ engine for realtime AI agents. SDKs for Node.js, Unity, & Unreal.Reliability and performanceNov 3, 2025 11mo11 months agoThe complete guide to measuring and optimizing TTS latencyLearn how to measure TTS latency accurately across providers. Understand how WebSocket, streaming, and audio post-processing affect latency, and try our interactive demo to benchmark latency across providers.Developer toolsOct 29, 2025 12mo12 months agoUnreal AI Runtime: The first unified interactive AI toolkit for game developersThe first unified toolkit for building realtime interactive AI experiences in Unreal. With 100+ models, built-in observability, and an intuitive visual editor.Developer toolsOct 22, 2025 1y1 year agoIntroducing Inworld CLIInworld CLI - a unified toolkit for building, deploying, and optimizing realtime conversational AI and voice agentsInfrastructureOct 15, 2025 1y1 year agoThe new AI infrastructure for scaling games, media, and charactersBuild scalable AI experiences and characters for games and media with Inworld Runtime. The AI infrastructure trusted by Xbox, Ubisoft, & NVIDIA.Reliability and performanceOct 1, 2025 1y1 year agoInworld + LiveKit: Unlocking studio-quality voice AI for real-time experiences at scaleCreate multiplayer games and AI agents with Inworld + LiveKit. Real-time voice AI, global edge infrastructure, accessible pricing. Build today.ResearchSep 25, 2025 1y1 year agoInworld meets Pipecat: Raising the bar for realtime voice AIStudio-quality voice AI with realtime TTS in your Pipecat pipeline.Reliability and performanceSep 15, 2025 1y1 year agoYour AI is boring: Boost engagement and drive immediate performance improvements with Inworld Runtime and custom Mistral AI modelsPartner with Mistral AI and Inworld to build consumer AI apps users love. Custom models, automated infrastructure, and instant A/B testing. Get started.Reliability and performanceSep 10, 2025 1y1 year agoWishroll / Status: Cutting AI costs by >95%, scaling to 500K+ DAUs, and driving time spent per user to over 1.5 hours per dayInworld helped Status achieve 1.5hr daily engagement, sub-second response times, and a top App Store ranking, at 95% lower AI cost. Read the case study.AI and agentsSep 3, 2025 1y1 year agoInworld + Vapi: Powering the next generation of expressive, real-time voice agentsBuild expressive voice agents with Inworld + Vapi. Test prompts, voices, and flows pre-production. Create assistants, coaches, and characters. Start today.Product and companyAug 15, 2025 1y1 year agoIntroducing Realtime TTSState-of-the-art voice AI for every developer. Studio quality at $15/1M characters, 200ms latency, free zero-shot cloning. Try the playground now.Product and companyAug 7, 2025 1y1 year agoInworld + NLX: Enabling multi-modal consumer experiences with SOTA voice AIBuild voice AI apps with Inworld + NLX. 200ms latency, 11 languages. Start building today.Product and companyAug 1, 2025 1y1 year agoHow we made state-of-the-art speech synthesis scalable with ModularVanilla vLLM couldn't meet Inworld's latency and cost targets. This is how a Modular MAX migration achieved sub-200ms streaming at production-scale pricing.ResearchAug 1, 2025 1y1 year agoHow Inworld helped an AI game with 20 million players reach profitabilityInworld built custom APIs, reduced Death by AI's costs, fixed latency issues, and enabled quality localization across languages. See the full case study. About this blog
Inworld's own blog on inworld.ai, found by the daily check.
Posts per week, last 26 weeks
1 a month over the last six months (or since the feed's oldest post)
History
Listed on Blogs Oct 13, 2026. No changes to its feed, status or cadence since.
Blogsby Ibis