NVIDIA's Nemotron 3.5 Super VL tops US open-weight models at 26 on Intelligence Index
ArtificialAnlys · x · 2026-10-15
Artificial Analysis benchmarks NVIDIA's new Nemotron 3.5 releases: Nemotron 3.5 Super VL scores 26 on the Intelligence Index, now the highest-scoring US open-weights model, edging out Thinking Machines' Inkling Small (26) and passing NVIDIA's previous flagship Nemotron 3 Ultra (23) and Meta's Muse Glimmer (17).
- Super VL delivers 2x the intelligence of Nemotron 3 Super (13) while using 60% fewer output tokens per task (46k vs 114k)
- Companion release Nemotron 3.5 Bolt 74B-A9B 24GB, first of the new Bolt family, scores 21 — 4th among sub-100B models and #1 among US models in that class
- The US still trails the UAE (K2 Horizon 375B A23B, 31) and China (MiMo-V2.6-Pro, 46) on open-weights intelligence
- On a pre-release DeepInfra NVFP4 endpoint, Super VL hits 630 tokens/s at 100k-token prompts — the fastest 20+ score model measured, with 1.8s first-token latency
Related event: NVIDIA Nemotron 3.5 Super VL tops US open-weight models(2 posts)→
More from Models
- NVIDIA's Nemotron 3.5 Super VL hits CoreWeave: 12B-active multimodal MoE at $0.25/M input tokens — wandb · 2026-10-16
- Opus 5 might be a very deep model or use looping, commentator speculates — scaling01 · 2026-10-16
- GPT-6.1 Sol audit finds 6% of arXiv papers had false main theorems, down to 1.5% recently — nitarshan · 2026-10-16
- Nvidia-backed post-training pushes Nemotron 3.5 VL past frontier on dermatology diagnosis — LoganGrasby · 2026-10-16
- Gemini 4 Argon scores 97.5% on ARC-AGI-3, second lab above 95% — patience_cave · 2026-10-16
- Microsoft MAI claims SOTA 3T-param in-house base models, post-training results coming — soldni · 2026-10-16