TEL AVIV, Israel — September 11, 2025 — Leads & Copy — Deepdub has launched Lightning 2.5, a voice AI model designed for real-time, production-grade voice experiences. The model supports multilingual AI agents, contact center automation, content localization, and personalized media and gaming.
Built on Deepdub’s language and speech foundation models and NVIDIA GPUs, Lightning 2.5 delivers 2.8X more throughput and 5X higher concurrency than the previous version, with latency as low as 200 milliseconds. According to Deepdub, the updated model preserves speaker identity, intonation, and emotional nuance, which makes it suitable for lifelike voice experiences.
Deepdub, a member of the NVIDIA Inception program, has optimized Lightning for deployment across NVIDIA GPUs, including NVIDIA H100, NVIDIA A100, NVIDIA L40S, and NVIDIA A10G. The company will showcase Lightning at IBC 2025 at the AI Pavilion, Hall 14, Booth 14.B53.
Contact: Sarah Schloss, Headline Media, sarah.schloss@headline.media, +1 914 506 5104
Source: Deepdub
