Tech Meridian← ALL MODELS
PROMY MERIDIAN

NVIDIA · MODEL RELEASE TRACKER

Nemotron 3 Ultra

Nemotron 3 Ultra is a named model in NVIDIA's Nemotron series. The supplied source reports that LangChain tuned its Deep Agents harness for Nemotron 3 Ultra and achieved near-frontier performance at approximately 10× lower cost per run; a tuned model profile is available on GitHub.

CURRENT SNAPSHOT3/5 DIMENSIONS WITH DATA

The dimensions that change the decision.

PRICING
Reported cost per run (relative)Reported to operate at approximately 10× lower cost per run (as stated in the source).
CONTEXT WINDOW

Not established from the available sources.

MODALITIES

Not established from the available sources.

BENCHMARKS
Concurrent-user capacity with NIM optimizationsFull‑stack NIM optimizations increased concurrent‑user capacity by 2.5× when serving Nemotron 3 Ultra (reported by the source).
Reported performanceLangChain tuning of its Deep Agents harness for Nemotron 3 Ultra was reported to achieve near-frontier performance.
AVAILABILITY
Tuned model profile availabilityA tuned model profile for Nemotron 3 Ultra is reported to be available on GitHub.

RELEASE TIMELINE

Published, source-backed release events only.

2 SOURCES · IMPORTANCE 8.0

Nvidia unveils Nemotron 3.5 Super VL and lightweight Bolt models for enterprise agents and local AI PCs

Nvidia announced new open-weight models aimed at always-on agents and edge/PC deployment: Nemotron 3.5 Super VL (120B parameters) for enterprise agents, plus two lightweight Bolt models including Nemotron 3.5 Bolt 74B-A9B 24GB and the smaller Bolt E4B for local AI PCs and workstations. Nvidia says Super VL and Bolt 74B-A9B are available now (Bolt E4B coming later this month), claims Super VL can match larger Nemotron 3 Ultra on instruction-following and outperform some competing models in post-trained agent tasks, and named early users such as Skydance, EY, Ramp, CrowdStrike, Perplexity and Salesforce.

→
2 SOURCES · IMPORTANCE 8.0

NVIDIA releases Nemotron 3.5 Super VL, a 120B-parameter open multimodal model for agents

NVIDIA announced Nemotron 3.5 Super VL, an open 120B-parameter mixture-of-experts multimodal model that activates 12B parameters per token to support agentic workflows combining reasoning, instruction following, tool use, and text, image, document, and video understanding. NVIDIA says it achieves up to 5× higher single‑GPU throughput on agentic tasks versus comparable open models, supports post‑training via NeMo (SFT, LoRA, RL) with resources on Hugging Face and a GitHub usage cookbook, and employs inference optimizations such as LiLiCorr speculative decoding and NVFP4 quantization; partners including Scale AI, Ramp, and AI21 report substantial cost, latency, or accuracy improvements in production workloads.

→

VERIFIABLE FACTS

Every value stays attached to a source and date.

WHAT CHANGED

Stored passport versions, without reconstructed history.

Passport updated7 facts
AVAILABILITY · Available on AWS− Listed among NVIDIA Nemotron models that are already available on AWS.
PRICING · Reported cost per run (relative)+ Reported to operate at approximately 10× lower cost per run (as stated in the source).
BENCHMARKS · Reported performance+ LangChain tuning of its Deep Agents harness for Nemotron 3 Ultra was reported to achieve near-frontier performance.
AVAILABILITY · Tuned model profile availability+ A tuned model profile for Nemotron 3 Ultra is reported to be available on GitHub.
Passport updated5 facts
AVAILABILITY · Available on AWS+ Listed among NVIDIA Nemotron models that are already available on AWS.
Passport updated4 facts
CAPABILITIES · Model size (parameters)Implied to be greater than 120 billion parameters — the article states Nemotron 3.5 Super VL (120B) 'uses fewer parameters than the Nemotron 3 Ultra.'→Reported to use more parameters than Nemotron 3.5 Super VL; Nemotron 3.5 Super VL is described as a 120-billion-parameter model, implying Nemotron 3 Ultra has more than 120 billion parameters.
Passport updated4 facts
CAPABILITIES · Parameter count+ Implied to be greater than 120 billion parameters — the article states Nemotron 3.5 Super VL (120B) 'uses fewer parameters than the Nemotron 3 Ultra.'
Passport updated3 facts
CAPABILITIES · Largest reasoning model / reasoning capability+ The sources describe Nemotron 3 Ultra as the Nemotron lineup's largest reasoning model and state it is intended to tackle tough reasoning tasks.
CAPABILITIES · Optimized for NVIDIA hardware+ The excerpts state that every Nemotron variant, which includes Ultra, is optimized for NVIDIA hardware.
Passport created1 facts