NVIDIA Open-Sources Nemotron 3.5 Super VL, a Cost-Effective Multimodal Agent Model
NVIDIA has released Nemotron 3.5 Super VL, an open-weights multimodal model designed for enterprise AI agents with significantly improved cost efficiency compared to its predecessor. Released on October 15, 2026, the 120-billion-parameter mixture-of-experts model activates only 12 billion parameters per token, balancing capability with computational efficiency. The model ranks among the top performers on Design Arena's Website leaderboard while reducing output token costs to just $0.50 per million tokens—approximately five times less expensive than Nemotron 3 Ultra's $2.50 per million output tokens.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.