NVIDIA's Nemotron 3.5 Transcribe ties for #1 open-weight STT at 2.8% WER with just 4B active params
ArtificialAnlys · x · 2026-10-28
Artificial Analysis benchmarked NVIDIA's new open-weight speech-to-text model Nemotron 3.5 Transcribe (30B total, 4B active, non-streaming): it scores 2.8% WER on the AA-WER Index, tying Mistral's Voxtral Small for #1 among open-weight models and beating NVIDIA's previous best Canary Qwen 2.5B (4.3%) and Thinking Machines' Inkling (3.5%).
- Efficiency: matches Voxtral Small's accuracy with 6x fewer active parameters (4B vs 24B); similar-sized open models sit around 4% WER.
- Voice-agent frontier: best open-weight score on AA-AgentTalk at 1.8% WER vs Voxtral Small's 2.3%.
- Speed & cost: on DeepInfra it hits a 75x Speed Factor and costs $2.00 per 1,000 minutes of audio.
Related event: NVIDIA Open-Sources Nemotron 3.5 Transcribe, Topping Open ASR Leaderboards(7 posts)→
More from Models
- Cryptographer points abliterated GLM at Telegram's secret-chat crypto — and it 'killed itself' — matthew_d_green · 2026-10-29
- Qoder offers DeepSeek-V4.1-Flash-1017 free through Oct 31 with zero credit cost — Aiden_Tech_Ai · 2026-10-29
- Top reasoning effort costs up to 42x more on tasks low effort already solves — zainhas · 2026-10-29
- RealityBench: 35 VLMs tested on physical work understanding, measuring stays a weak spot — arnabdotorg · 2026-10-28
- Permissive flagship open-model licenses fell from 75% to 35% in months — bookwormengr · 2026-10-28
- Anthropic Gets 80% of Revenue From API Pricing; DeepSeek Costs 20x Less — bindureddy · 2026-10-28