NVIDIA released Nemotron 3.5 Super VL, an open 120-billion-parameter model for specialized agents, alongside two Bolt edge models.

NVIDIA released Nemotron 3.5 Super VL as an open 120-billion-parameter model for specialized, always-on agents. Its developer announcement emphasizes multimodal reasoning and enterprise workflows involving document reading and checking information across sources.

NVIDIA describes the model as delivering the fastest task completion and highest accuracy for specialized always-on agents. These are vendor claims: the supplied excerpts do not provide benchmark scores, comparison systems, or test conditions sufficient to independently evaluate them.

The accompanying edge releases are Nemotron 3.5 Bolt 74B-A9B 24GB and Nemotron 3.5 Bolt E4B 4GB. NVIDIA explicitly describes the former as balancing intelligence and speed per gigabyte for local agentic workloads; the supplied excerpt does not fully describe the latter’s optimization target.

The Bolt names here refer to NVIDIA models, not Bolt Graphics, the separate chip-design startup also mentioned in the day’s sources. In a related software announcement, Adobe said a brand-concierge service powered by NVIDIA Nemotron was planned for enterprise customers later in the year, but the source did not identify a specific Nemotron version.