Blog
Research and engineering notes on realtime multimodal inference.
-
Nemotron-VoiceChat: 64 Tool-Using S2S Agents on 1 GPU
-
Nari TTS and STT Are Now Available Through Speko Router
-
Input Streaming for Nari Qwen3-TTS
-
Word Timestamps for Qwen3-ASR without the Slowdown
-
Nari Model APIs Are Now Generally Available
-
PersonaPlex 7B: A Full-Duplex Conversation Below 10 Cents an Hour
-
Nari Labs Leads Coval’s Voice AI Benchmarks
-
Making Pyannote Ultrafast: 100+ Hours of Audio in Under 100 Seconds
-
Expanding the Pareto Frontier of STT with Nari Qwen3-ASR
-
Introducing Nari Qwen3-TTS
-
Research
Nari Qwen3-TTS Preview Ranked #1 in Latency on Speko
-
Research
Pushing the Speed-Cost Frontier for Qwen3-TTS
