AI Model Release Tracker

Nemotron 3.5 Super VLOpen Weight

Released

Nemotron 3.5 Super VL is an AI model released by NVIDIA on Thursday, Oct 15 2026, 65 days after Nemotron 3.5 Lightning. It is an open-weight model — the trained weights are available to download and run. It is a 120B parameter model with a 1M token context window. Benchmark results (shown below) cover SWE-Bench Pro, SWE-Bench Verified, SWE-Bench Multilingual, LiveCodeBench, Terminal-Bench 2.1, BrowseComp, and 2 more.

Benchmarks

Coding

SWE-Bench ProAgentic coding — Can the AI fix real bugs in real software? It's handed actual problems from open-source projects and has to write code that genuinely solves them. Higher is better.
51.78%
#20 of 24
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better.
71.1%
#38 of 59
SWE-Bench MultilingualMultilingual coding — Like SWE-Bench, but the coding problems span many programming languages, not just one. Tests how broadly the AI can code. Higher is better.
67.72%
#13 of 15
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better.
88.22%
#8 of 15

Terminal & CLI

Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better.
66.7%
#29 of 35

Agentic & tool use

BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better.
78%
#21 of 30

Reasoning & science

GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better.
89.52%
#19 of 61
IFBenchInstruction following — Tests whether the AI can follow detailed instructions and satisfy multiple constraints at once. Higher is better.
70.86%
#3 of 3

Compare Nemotron 3.5 Super VL with

Nemotron 3.5 Super VL

Suggested comparisons

Nemotron 3.5 Super VLvsNemotron 3.5 Bolt 74BNemotron 3.5 Super VLvsGPT-6.1 SolNemotron 3.5 Super VLvsClaude Haiku 5.5Nemotron 3.5 Super VLvsGemini 4 ArgonNemotron 3.5 Super VLvsMuse Spark OpenNemotron 3.5 Super VLvsGrok 4.7Nemotron 3.5 Super VLvsDeepSeek-V4.1-FlashNemotron 3.5 Super VLvsMistral Large 4Nemotron 3.5 Super VLvsKimi K3Nemotron 3.5 Super VLvsGLM-5.3-FlashNemotron 3.5 Super VLvsQwen3.8-Max-0902

About

Nemotron 3.5 Super VL, released October 15, 2026, extended the Super line to agents working with images, documents and video alongside text. NVIDIA paired a C-RADIO vision encoder with its hybrid Mamba-Transformer mixture-of-experts backbone, retaining 120 billion total parameters, 12 billion active parameters and up to 1 million tokens of context.

NVIDIA reported launch scores of 89.52% on GPQA Diamond and 71.10% on SWE-Bench Verified for the BF16 checkpoint. It published BF16 weights for customization and an NVFP4 variant for deployment, both under OpenMDW-1.1. The release was open weight; its training-data inventory also listed private datasets.

Frequently asked questions

Nemotron 3.5 Super VL was released by NVIDIA on Thursday, Oct 15 2026.

All NVIDIA releases

11 tracked