Skip to main content

New LLMs, Speech Providers & Speed Gains ๐Ÿ”ฅ

Google CX Studio, Mistral, Cartesia & Latency Optimizations

Written by Tomer Yair Zemel
Updated this week

Hey there Live Hub Legends!

At Live Hub, we give you the flexibility to build advanced, natural voice AI using the providers and models that best fit your needs. The strongest voice experiences happen when AI models, speech technologies, and real-world voice infrastructure work together seamlessly.

This release strengthens that foundation. Weโ€™ve added new integrations, expanded our LLM and speech ecosystem, and implemented key latency optimizations, helping your agents respond faster and perform more naturally in production environments.


๐Ÿค– New!! Google CX Studio Integration

Live Hub now integrates with Google CX Studio, enabling deployment of Gemini-powered generative agents alongside existing Dialogflow CX workflows.

What this unlocks:

  • Build voice bots on Googleโ€™s next-generation conversational platform

  • Seamlessly connect Google bots to any voice channel

  • Integrate easily with existing contact center environments

Shape

๐Ÿ”Š Cartesia TTS + Early Access Speech Providers

Elevate your botโ€™s voice with Cartesia TTS (via Bring Your Own Key). Cartesia adds high-fidelity, expressive speech output for a more natural, human-like experience.

We are also preparing additional integrations for:

  • Soniox

  • Speechmatics

  • ElevenLabs (STT)

  • Inworld (TTS)

Early access is available. If youโ€™re interested in testing these providers, please contact our support team.


๐Ÿง Live Hub AI Agents: Expanded LLM Support - You Choose the Brain

Weโ€™ve expanded LLM options to give you greater control over the intelligence layer behind your agents:

  • Google Vertex AI โ€“ Gemini 3 Flash and Pro via Vertex AI

  • Mistral โ€“ High-efficiency, performance-optimized models

  • OpenAI โ€“ GPT-5.2

  • Anthropic โ€“ Claude 4.5 (Sonnet & Haiku)

  • Groq โ€“ Ultra-fast inference with Llama-4 Maverick & Scout

What this unlocks:

  • Select the optimal AI engine per use case

  • Balance reasoning power, latency, and cost

  • Avoid single-vendor lock-in

You choose the intelligence layer & Live Hub handles real-world voice deployment.


โšก AI Agents Real-Time Performance Optimizations

To support production-grade conversational AI, weโ€™ve introduced infrastructure upgrades focused on speed and conversational flow:

  • TTS Streaming โ€“ Reduces time-to-first-audio

  • Active Listening Mode โ€“ Smoother conversational turn-taking

  • Zero-Freeze VAD โ€“ Mitigates intermittent speech flow interruptions

What this unlocks:

  • Faster, more natural responses

  • Reduced awkward pauses

  • Improved customer perception and engagement

For customer-facing AI agents, these improvements directly impact satisfaction and perceived intelligence.


โœ๏ธ Account Rename Is Now Possible

It's finally here! You can now rename your account directly from the Settings page, simplifying multi-account management for partners and enterprise teams.


Ready to Deploy Production-Grade Voice AI?

Live Hub continues to evolve as a production-ready voice AI platform, combining AI flexibility with real-world telephony reliability.

Interested in early access to upcoming speech providers?
Reach out to us anytime and anywhere, and our support team will assist.

๐Ÿ’ก Got an amazing idea?


Weโ€™d love to hear from you - your feedback shapes the future of Live Hub.
๐Ÿ“ง Submit Feedback


๐Ÿ‘‹ Need further directions?


Click here to schedule a session with your CS team manager.


We appreciate your continued support of Live Hub and look forward to hearing what you build next.

Happy Connecting! ๐ŸŽ‰


The Live Hub Team ๐ŸŒ


โ€‹by AudioCodes

Did this answer your question?