Follow-up: NVIDIA teases new model ahead of possible GTC reveal
testingcatalog · x · 2026-10-15
testingcatalog posted a link as another teaser for NVIDIA's upcoming model, pointing back to its earlier report that Nemotron 3.5 Super VL may arrive at GTC on Wednesday the 21st.
The model is rumored to use a hybrid Mamba-Transformer MoE architecture (120B total / 12B active) with image and video input; related documentation pages have been pulled by NVIDIA. See the original post for full details.
More from Models
- Open-weights TYGER model combines ASR with full-duplex turn control, 250ms latency — KokaOP · 2026-10-15
- Step 5 Preview architecture blog details Sparse GQA with block-wise token merging cutting indexer cost 8x — teortaxesTex · 2026-10-15
- Fish Audio Drama 3 lands in vLLM-Omni: 128 concurrent streams on one H200 — vllm_project · 2026-10-15
- Why is GPT-6 Astra ranked so low on LM Arena? Outscored by GPT-5.6 and 6.1 — AreaExact7824 · 2026-10-15
- Stance test heat map shows Sonnet 5.5 resists framing manipulation other models succumb to — teortaxesTex · 2026-10-15
- Argon (Medium) burned 85% of a 5-hour usage limit in just 5 minutes — Avalunne · 2026-10-15