Skip to main content

text model comparison

Nemotron 3.5 Lightning vs Nemotron 3.5 Super VL

Nemotron 3.5 Lightning and Nemotron 3.5 Super VL are both current text models, from NVIDIA. On Arena Text, Nemotron 3.5 Lightning scores 1353 Elo and Nemotron 3.5 Super VL 1412 Elo (Arena, measured October 28, 2026). Both accept 1,000,000 tokens of context.

Higher Arena Text
Nemotron 3.5 Super VL1412 Elo against 1353 Elo
Newer release
Nemotron 3.5 Super VLOct 2026 against Aug 2026

Data verified November 3, 2026 · Dataset 2026-11-03

Share this

Side by side

Nemotron 3.5 Lightning and Nemotron 3.5 Super VL, line by line

Prices in effect today, compared at the same tier wherever both list one. The better value in each row is highlighted.

FactNemotron 3.5 LightningNemotron 3.5 Super VL
Independent benchmarks
Arena Text1353 Elo (#49 of 51)1412 Elo (#45 of 51) (better)
Arena WebDevNot measured1399 Elo (#41 of 45)
Specs
ProviderNVIDIANVIDIA
Release dateAugust 11, 2026October 15, 2026
Accessopen weightsopen weights
Context window1,000,000 tokens1,000,000 tokens
Max outputNot reportedNot reported
Inputtexttext, image, video
Outputtexttext

Scores:Arena · 2026-11-03

Open both in the interactive comparison

Charts

Against every current text model

Where Nemotron 3.5 Lightning and Nemotron 3.5 Super VL land among the current models of their task, on every independent benchmark, on price and on context.

Where each stands

On Arena Text, Nemotron 3.5 Lightning ranks #49 and Nemotron 3.5 Super VL #45 of 51 current models. Each bar is the share of current models it beats: on score, price and context.

Nemotron 3.5 Lightning and Nemotron 3.5 Super VL among the current text models, one row per benchmark, price and context window.

Independent benchmarks

Arena Text

Nemotron 3.5 Lightning:#49of 51

1,353 Elo · ahead of 4%

Nemotron 3.5 Super VL:#45of 51

1,412 Elo · ahead of 12%

Arena WebDev

Nemotron 3.5 Lightning: Not scored

Nemotron 3.5 Super VL:#41of 45

1,399 Elo · ahead of 9%

Price and context

Context window

Nemotron 3.5 Lightning:#26of 64

1M tokens · longer than 22%

Nemotron 3.5 Super VL:#26of 64

1M tokens · longer than 22%

Verdict

When to choose each

Nemotron 3.5 Lightning

Where it leads

No measured lead over Nemotron 3.5 Super VL in this data.

Best for

  • Fast, low-cost agent subtasks such as tool calls
  • Self-hosted agents on one GPU

Watch out for

  • On NVIDIA's own table it trails Qwen 3.6 35B A3B on SWE-bench Verified (51.56 against 70.12)
  • NVIDIA sells no per-token API for it; hosted prices come from third parties

Nemotron 3.5 Super VL

Where it leads

  • Higher Arena Text score, 1412 Elo vs 1353 Elo
  • Accepts image and video input

Best for

  • Agents that read screenshots, charts and documents
  • Self-hosted multimodal agents on NVIDIA hardware

Watch out for

  • NVIDIA calls it 5x faster than the next best open model in its class; that speed claim is not measured independently here
  • NVIDIA sells no per-token API for it, so hosted prices come from third parties

From benchmark to production

Building on Nemotron 3.5 Lightning or Nemotron 3.5 Super VL?

Public scores rarely hold on your own data. We test both models on your real tasks, then build the agent or pipeline around the one that wins.

AI Model Cost AuditLaunch price $100, regular price $150What the audit includes

FAQ

Nemotron 3.5 Lightning vs Nemotron 3.5 Super VL FAQ

Which scores higher on independent benchmarks, Nemotron 3.5 Lightning or Nemotron 3.5 Super VL?

Nemotron 3.5 Super VL leads on Arena Text, 1412 Elo to 1353 Elo. Scores are independent evaluations, not either vendor's launch figures.

Which has the larger context window, Nemotron 3.5 Lightning or Nemotron 3.5 Super VL?

Both accept 1,000,000 tokens of context.

How do Nemotron 3.5 Lightning and Nemotron 3.5 Super VL differ in features?

Nemotron 3.5 Lightning takes text input; Nemotron 3.5 Super VL takes text, image and video.

Production evaluation

Choosing between Nemotron 3.5 Lightning and Nemotron 3.5 Super VL?

Public benchmarks narrow the shortlist. We run both on your own tasks, data and cost limits, then build the routing and review controls to run the winner in production.

  • A head-to-head evaluation on your real workload
  • Routing between models by cost and difficulty