NVIDIA releases Nemotron 3.5 Super VL and Nemotron 3.5 Bolt 74B-A9B

Artificial Analysis says NVIDIA's Nemotron 3.5 Super VL scores 26 on its Intelligence Index, making it the leading US open-weight model, and roughly doubles its predecessor's score.

It uses about 60% fewer output tokens per task and runs at about 630 tokens per second on a pre-release DeepInfra endpoint.

The Nemotron 3.5 Bolt 74B-A9B scores 21, and Artificial Analysis notes that the US still trails the UAE and China on open-weight intelligence.