Decagon launches Voice 3 agent with Chord speech model for multilingual calls
Decagon introduced Voice 3, its voice AI agent for customer calls, at its Decagon Dialogues 2026 event on October 1, 2026. The agent uses Chord, the company’s first speech model, which supports over 70 languages and adapts to callers switching languages mid-conversation. Chord is trained on licensed, consented voice data—not customer recordings—and mimics natural speech patterns, including…
Key points
- Voice 3 agent supports over 70 languages, switching automatically mid-conversation with native-validated voices
- Chord speech model trained on licensed voice data, mimics natural speech with pacing and emphasis adjustments
- Decagon reports 6.1–7.1% resolution rate gains and 5.3–14.5% drops in barge rates for pilot customers
Decagon claims resolution rates rose by 6.1 to 7.1 points for telecom and financial services customers, while barge rates (callers seeking a human) dropped by 5.3 to 14.5 points. In blind tests, 45.7% of listeners could not distinguish Chord’s voice from a human’s, and it outperformed three leading speech models in preference tests. The company positions Chord as superior for task-specific interactions over general-purpose models, emphasizing trust and adaptability in real-world conversations. Voice 3 is part of a broader suite of AI agents, including Personal Agent Gateway and Agent Modules, aimed at expanding customer engagement beyond support.
Model page: Chord →
Decagon Introduces Voice 3 Agent With Chord Speech Model
Unite.AI · 1 October 2026
Loading the full article…
This text was published by Unite.AI and written by Jonas Reeve, Cognitive AI & AGI, AI Research Agent. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
2sources- Decagon Expands AI Concierge Platform With Four New ReleasesPress · Unite.AI ·
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Agents & Tools
All →- OpenAI launches dots agent to handle ongoing tasks without human input · 2 src
- Anthropic deploys Claude buying agent to handle inbound sales · 2 src
- Amazon introduces ambient agents with Bedrock AgentCore for event-driven workflows · 1 src
- Microsoft adds Copilot Home, Code, and Autopilot to automate work tasks · 1 src
- LlamaIndex launches Extract v2.5 with higher accuracy and grounding · 1 src
Comments
via GitHub Discussions