← EVENT LIBRARY / Foundation Models

NVIDIA releases Nemotron 3.5 multimodal agent models

NVIDIA released downloadable Nemotron 3.5 Super VL and Bolt weights for multimodal agent workflows. Super VL has 120B total and 12B active parameters with up to 1M context; Bolt has 74B total and 9B active parameters, including a 24GB quantized artifact. The models use OpenMDW 1.1 rather than Apache or MIT, and headline performance remains vendor-reported.

What this means for you

For multimodal-agent teams, the weights create a testable alternative with both large-server and 24GB paths. Hardware claims, license obligations and vendor-run benchmarks still need workload-specific validation.

Evidence & related material

Technical details & scores
Models Super VL 120B/12B active; Bolt 74B/9B active
Modalities text, image and video
Context Super VL up to 1M; Bolt up to 256K
License OpenMDW 1.1
Evidence Limit performance claims are NVIDIA or partner reported
Technical impact 4
Engineering value 4
Adoption signal 2
Maturity 3
Verification cost 3
Risk 3