NVIDIAReleased Oct 15, 2026

Nemotron 3.5 Super VL

120B/12B-active Mamba-2 hybrid with a vision encoder; native speculative decoding and switchable thinking.

Context window1M
Parameters120B
Provider endpoints0
Benchmark results19
Measured performance

Benchmark results

All leaderboards ↗
998.76

Effort not specified

Independent result
1,162.41

Effort not specified

Independent result
AutomationBench-AA ↗AA Index 4.3.2 collection
14.82%

Effort not specified

Independent result
SciCode ↗AA Index 4.3.2 collection
47.8%AA · Under review

Effort not specified

Independent result
GDP.pdf ↗AA Index 4.3.2 collection
8%

Effort not specified

Independent result
CritPt ↗AA Index 4.3.2 collection
10.57%AA · Under review

Effort not specified

Independent result
AA-Omniscience ↗AA Index 4.3.2 collection
-4.22

Effort not specified

Independent result
69%

Effort not specified

Independent result
LiveBench Overall ↗2.0 · 2026-10-14 (dataset 2026-09-02)
51.47%

Effort: max · single run · equal mean of 7 suites

Independent result
LiveBench Design ↗2.0 · 2026-10-14 (dataset 2026-09-02)
60.7%

Effort: max · single run

Independent result
LiveBench Research ↗2.0 · 2026-10-14 (dataset 2026-09-02)
70.9%

Effort: max · single run

Independent result
LiveBench Browser Use ↗2.0 · 2026-10-14 (dataset 2026-09-02)
4%

Effort: max · single run

Independent result
LiveBench Coding ↗2.0 · 2026-10-14 (dataset 2026-09-02)
65.5%

Effort: max · single run

Independent result
LiveBench Agentic ↗2.0 · 2026-10-14 (dataset 2026-09-02)
15.8%

Effort: max · single run

Independent result
LiveBench Reasoning ↗2.0 · 2026-10-14 (dataset 2026-09-02)
78.8%

Effort: max · single run

Independent result
Access & cost

Prices across providers

0 quotes

Model specifications

Developer
NVIDIA
Modality
Language
Context
1,000,000 tokens
Maximum output
Not reported
Weights
Open weights
Released
Oct 15, 2026
First discovered
Oct 16, 2026

Exact identities

  • artificial-analysise2c0623e-c855-4a61-bb66-be1f1ed43234
  • livebenchnemotron-super-vl-3.5

Sources

Updated Oct 21, 2026

More from NVIDIA

View all ↗