From the NVIDIA edition of October 15, 2026
Developer blog: Nemotron 3.5 Super VL MoE activates 12B parameters; NeMo post-training and NIM for production
NVIDIA · Technology vendor / AI platforms · Global
- Publisher
- NVIDIA Technical Blog
- Original publication
- 2026-10-15
- Source retrieved
- 2026-10-16
- Event date
- 2026-10-15
What happened
The Oct 15 technical blog details Nemotron 3.5 Super VL as a 120B-parameter MoE that activates 12B parameters per token, combining reasoning, instruction following, tool use and text/image/document/video understanding. Vendor-attributed partner results include Scale AI for Skydance vendor-credit processing (claimed 40x lower inference cost and 5x lower latency), Ramp receipt-to-JSON (up to 50% higher throughput and 10x lower latency), and AI21/Alice cybersecurity post-training (41% accuracy improvement). Hermes Bench 63.44% and LangChain evals 87.05% are reported versus Nemotron 3 Ultra. Enterprise support for NIM is through NVIDIA AI Enterprise. Training notes include Indic data with IIT Madras/AI4Bharat and open datasets on Hugging Face.
Why it matters
Useful engineering detail for SLED platform teams sizing MoE inference and post-training; partner metrics remain vendor-attributed commercial examples, not public-sector evaluations.
Evidence and measured results
Primary NVIDIA technical blog Oct 15, 2026 by Chris Alexiuk and Chintan Patel; includes tables and partner figures without independent protocols.
Limitations and uncertainty
Partner cost/latency/accuracy figures lack published task definitions and sample sizes in this page; AI-generated summary banner notes possible incompleteness.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-16; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Buyers need credible sizing and customization paths for multimodal agents without treating vendor partner charts as guaranteed savings.
- Stakeholders
- Enterprise architects
- AI platform leads
- Finance for license TCO
- Security
- Discovery questions
- Which partner-style workflow is closest to ours?
- What baseline cost/latency do we measure today?
- Is AI Enterprise already entitled on our GPUs?
- Value hypothesis
- Interpretation: MoE efficiency plus post-training may improve specialized agents; this source's partner percentages are vendor-attributed.
- Potential engagement
- Architecture workshop plus a single post-train PoC scoped to one document or security workflow with buyer-owned metrics.
- Unsupported claims
- That base Super VL beats Claude/GPT generally
- Guaranteed 40x inference cost reduction
- That Hermes/LangChain scores transfer to agency tasks
Pre-sales engineering
Role takeaway
- Fit
- Strong for NVIDIA-centric agent platforms needing multimodal MoE; weaker where CUDA stack is unavailable.
- Architecture and integration
- NeMo post-train then open or NIM serve with optional Dynamo multi-node; NVFP4 on Blackwell with fallback paths noted in ecosystem recipes.
- Prerequisites
- NeMo tooling
- Labeled or preference data
- Eval harness
- GPU operators
- Constraints
- Partner metrics lack protocols here
- Indic/training details are research context, not SLED requirements
- NIM support needs AI Enterprise
- Security
- Tool-using agents need sandboxing and audit; cybersecurity post-train example is not a control baseline.
- Proof of value
- Reproduce one partner-like task on agency data
- Report active-parameter memory footprint
- Compare NIM vs open serving support model
Delivery
Role takeaway
- Work
- Build post-train pipeline, run controlled evals, document deployment mode and support entitlement.
- Dependencies
- Data labeling capacity
- GPU cluster
- NVIDIA support channel if NIM
- App owners for tool APIs
- Ownership
- AI platform engineering with application owner for workflow acceptance.
- Skills and adoption
- ML engineers for NeMo RL/SFT; platform SRE for NIM/Dynamo.
- Governance checkpoints
- Training-data approval
- Eval protocol signed
- Production support entitlement confirmed
- Acceptance criteria
- Proposed: documented before/after on buyer task with n at or above agreed sample
- Proposed: chosen serving mode with rollback
- Risks
- Overfitting to narrow demos
- Support gap if open serving assumed free forever
- Eval leakage from vendor charts
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
MoE 12B-active inference; NeMo AutoModel SFT/LoRA and NeMo RL/Gym; NIM packaging; Dynamo KV-aware routing; NVFP4 quantization and LiLiCorr speculative decoding on DGX Spark.
Governance
Who approves, reviews and stays accountable for outcomes?
Post-training on org data/tools/policies is the stated path to specialized agents; buyers must govern training data rights and model change control.
Security and privacy
What data, permissions and controls need testing?
Cybersecurity partner example is attack-resilience post-training; still not a substitute for agency red-team and data-handling policy.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Open cookbooks and datasets lower barrier for research labs; production still needs MLOps skills.
Procurement
What should contracts, pricing and exit terms secure?
NIM enterprise support explicitly routed through NVIDIA AI Enterprise licensing.
Operating model
Which teams own the service once it runs?
Specialized agents require ongoing post-training and evaluation loops, not one-time model download.
Publication history
- 2026-10-15NVIDIA · Issue 407 resources
Stable resource ID: nvidia-nemotron-35-super-vl-devblog-20261015