What happened: Introducing NVIDIA Nemotron 3.5 ASR Streaming

NVIDIA Nemotron 3.5 ASR streaming models are now available on Baseten, offering low-latency speech recognition in English and 40 language variants.

Why it matters: Introducing NVIDIA Nemotron 3.5 ASR Streaming

Matters for engineers building real-time voice applications, live captioning, or interactive AI agents requiring sub-150ms transcription latency at scale.

Watch out: Introducing NVIDIA Nemotron 3.5 ASR Streaming

Models are 600M parameters each; performance benchmarks shown are single H100 results and may not generalize to other hardware or production conditions.

In the news

Introducing NVIDIA Nemotron 3.5 ASR Streaming

Baseten · Published August 4, 2026 · 3 min read

In 30 seconds

What happened: NVIDIA Nemotron 3.5 ASR streaming models are now available on Baseten, offering low-latency speech recognition in English and 40 language variants.
Why it matters: Matters for engineers building real-time voice applications, live captioning, or interactive AI agents requiring sub-150ms transcription latency at scale.
Watch out: Models are 600M parameters each; performance benchmarks shown are single H100 results and may not generalize to other hardware or production conditions.

Listen to this summary

Read the full article

nemotron

The patterns behind this

Each one covers how the technique works, when it earns its cost, and where it breaks.

The Agent Architect

One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.

Weekly email, one-click unsubscribe. We only use your address to send the briefing.