BenchLeader

Nemotron 3.5 Super VL vs Beam 1.0

Full profiles: Nemotron 3.5 Super VL (NVIDIA) and Beam 1.0 (Reflection): every benchmark, price and rank.

Updated · Data from 7 sources

As of 16 Nov 2026, Nemotron 3.5 Super VL and Beam 1.0 are neck and neck on quality (45.8 vs 45.9 on the BenchLeader Index, inside the ±9.7 margin), and Nemotron 3.5 Super VL is 2.2× cheaper at $0.275 vs $0.600 per 1M tokens blended. Beam 1.0 leads in every category we measure both in.

Key facts (as of 16 Nov 2026)

  • Nemotron 3.5 Super VL scores 45.8 on the BenchLeader Index and Beam 1.0 scores 45.9 (BenchLeader, 16 Nov 2026).
  • Nemotron 3.5 Super VL costs $0.275 per 1M tokens blended (list price, 16 Nov 2026).
  • Beam 1.0 costs $0.600 per 1M tokens blended (list price, 16 Nov 2026).
  • Of the 21 benchmarks both models have, Beam 1.0 leads 16 (BenchLeader, 16 Nov 2026).

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingBeam 1.037 vs 34 category score; FrontierCode 2.0 (AA) 5.5% vs 6.1%Medium
ReasoningBeam 1.042 vs 37 category score; HLE-Diamond (AA) 12.8% vs 8.8%Medium
KnowledgeBeam 1.046 vs 43 category score; AA-Omniscience v1.1 2.3 vs -2.9Medium
Long documentsBeam 1.064 vs 59 category score; AA-LCR v1.1 78.0% vs 69.0%Medium
Lowest costNemotron 3.5 Super VL$0.275 vs $0.600 per 1M blended; agentic coding session $0.014 vs $0.029–
Fastest first answerBeam 1.05.4 s (Beam 1.0) vs 13.7 s (Nemotron 3.5 Super VL) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • Nemotron 3.5 Super VL and Beam 1.0 are neck and neck on quality (45.8 vs 45.9, within the ±9.7 margin).
  • Beam 1.0 is stronger in coding, knowledge, long context, reasoning.
  • Nemotron 3.5 Super VL is 2.2× cheaper ($0.275 vs $0.600 per 1M blended).

At a glance

SpecNemotron 3.5 Super VLBeam 1.0
DeveloperNVIDIAReflection
Released15 Oct 20265 Oct 2026
Context window (tokens)1,048,5761,048,576
Compared atNemotron 3.5 Super VLBeam 1.0
Open weightsYesYes
Price in / out / cached input per 1M$0.200 / $0.500 / $0.020$0.400 / $1.20 / –

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryNemotron 3.5 Super VLBeam 1.0Leader
Reasoning36.542.3Beam 1.0
Coding34.436.9Beam 1.0
Agents & tools–41.3not comparable
Maths–38.8not comparable
Knowledge43.246.0Beam 1.0
Long context58.963.6Beam 1.0

On one chart

Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.

2550 avg75ReasoningCodingKnowledgeLong contextNemotron 3.5 Super VL, Reasoning: 36.5Nemotron 3.5 Super VL, Coding: 34.4Nemotron 3.5 Super VL, Knowledge: 43.2Nemotron 3.5 Super VL, Long context: 58.9Beam 1.0, Reasoning: 42.3Beam 1.0, Coding: 36.9Beam 1.0, Knowledge: 46.0Beam 1.0, Long context: 63.6

Not drawn, measured for fewer than two of these models: Agents & tools, Maths, Instruction following, Multimodal, Human preference.

Model, click to add or removeIndex

Where sources disagree

Across the 21 benchmarks both have, Nemotron 3.5 Super VL leads 5, Beam 1.0 leads 16. Every overall index that rates both puts Beam 1.0 ahead.

Overall indexNemotron 3.5 Super VLBeam 1.0Leader
BenchLeader Index45.845.9neck and neck
AA Intelligence Index v5.0.113.617.7Beam 1.0

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.

MeasureNemotron 3.5 Super VLBeam 1.0
Blended price per 1M tokens$0.275$0.600
Input / output per 1M$0.200 / $0.500$0.400 / $1.20
Cached input per 1M$0.020–
Price per task: chat reply$0.0002$0.0005
Price per task: summarize a 30-page report$0.0027$0.0055
Price per task: code edit$0.0020$0.0042
Price per task: agentic coding session$0.014$0.029
Price per task: structured extraction$0.0005$0.0010
Agentic coding session, input mostly cached$0.0059–
Cost per benchmark task (AA Index)$0.557$0.284
Output tokens per task (AA Index)81k34k
Of which reasoning56k22k
Time per task (AA Index)5 min119 s
Output speed271 tok/s288 tok/s
Time to first answer13.7 s5.4 s
Time to a complete answer, last measured 13 Nov 202636.4 s16.4 s

Per token is not per task: an average task of Artificial Analysis' Intelligence Index cost $0.557 on Nemotron 3.5 Super VL and $0.284 on Beam 1.0, and Nemotron 3.5 Super VL wrote 2.4× as many output tokens for the same work, which a per-token price hides.

Each model is compared at its best-ranked reasoning-effort setting. Nemotron 3.5 Super VL scores 45.8 and takes 14 s to its first answer token. Beam 1.0 scores 45.9 and takes 5 s to its first answer token. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

58 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherNemotron 3.5 Super VLBeam 1.0
BenchLeader IndexBenchLeader45.845.9
Coding scoreBenchLeader34.436.9
Knowledge scoreBenchLeader43.246.0
Long context scoreBenchLeader58.963.6
Reasoning scoreBenchLeader36.542.3
Agents & tools scoreBenchLeader–41.3
Maths scoreBenchLeader–38.8
Blended price $/MBenchLeader$0.275$0.600
Output speedBenchLeader271 tok/s288 tok/s
Time to first answerBenchLeader13.7 s5.4 s
Context windowBenchLeader1,048,576 tokens1,048,576 tokens
Reasoning
Humanity's Last Exam (AA)Artificial Analysis30.7%34.7%
HLE-Diamond (AA)Artificial Analysis8.8%12.8%
ARC-AGI-1ARC Prize2.0%†45.0%†
ARC-AGI-2ARC Prize0.8%†2.8%†
CritPtArtificial Analysis10.6%13.1%
CritPt v2 hard subset (AA)Artificial Analysis13.5%26.5%
MysteryMechanismVals AI–22.1%
Coding
SciCodeSciCode–48.5%
SciCode (AA)Artificial Analysis47.8%48.5%
IOIVals AI–22.8%
Code Migration v1.1Vals AI–22.9%
Vibe Code Bench v1.1Vals AI–66.4%
SWE Atlas v1.1Scale AI SEAL24.8%–
SWE Atlas v1.1: Codebase QnAScale AI SEAL26.6%–
SWE Atlas v1.1: Test WritingScale AI SEAL27.8%–
SWE Atlas v1.1: RefactoringScale AI SEAL20.0%–
ALE-BenchEpoch AI Benchmarking Hub–1097.5†
Terminal-Bench 4.0 (AA)Artificial Analysis0.0%3.0%
Design Arena WebsiteDesign Arena11601239†
FrontierCode 1.1 (AA)Artificial Analysis25.5%24.3%
FrontierCode 2.0 (AA)Artificial Analysis6.1%5.5%
Agents & tools
Terminal-BenchTerminal-Bench0.3%†6.4%†
APEX-AgentsMercor–48.8%†
GDPval-AA v2.1, normalised scoreArtificial Analysis33.1%33.9%
GDPval-AA v2.1Artificial Analysis11621177
Show all 58 benchmarks
MetricPublisherNemotron 3.5 Super VLBeam 1.0
Agents & tools
Finance Agent v2.1Vals AI–51.8%
Harvey's Legal Agent Benchmark v1.1Vals AI–4.6%
Legal Research BenchVals AI–28.9%
Tax Agent BenchVals AI–8.8%
Terminal-Bench 4.0 (Vals)Vals AI–4.0%
Terminal-Bench ScienceVals AI–1.4%
AutomationBenchArtificial Analysis14.8%48.5%
GDP.pdf (AA)Artificial Analysis8.0%11.6%
AA-Briefcase v1.1Artificial Analysis999976
Terminal-Bench-Science 0.1 (AA)Artificial Analysis0.0%0.5%
CWE-Bench (AA)Artificial Analysis–49.2%
DeepSec-Bench (AA)Artificial Analysis–6.9%
CyberGym end-to-end (AA)Artificial Analysis–45.0%
Maths
ProofBenchVals AI–18.0%
Knowledge
AA-Omniscience v1.1Artificial Analysis-2.92.3
BioMysteryBenchVals AI–42.6%
Excel Modeling Benchmark v1.1Vals AI–64.7%
AA-Omniscience v1.1: accuracyArtificial Analysis21.2%30.7%
AA-Omniscience v1.1: non-hallucinationArtificial Analysis69.5%59.0%
Multimodal
Humanity's Sixth SenseScale AI SEAL24.7%†–
Code2Video-AAArtificial Analysis–675
Long context
AA-LCR v1.1Artificial Analysis69.0%78.0%
MLCRArtificial Analysis13.3%17.8%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub–150.9
AA Intelligence Index v5.0.1Artificial Analysis13.617.7
AA Intelligence Index v4.3.2Artificial Analysis26.530.3
AA Progress Index v1Artificial Analysis283.9296.2
Vals IndexVals AI–43.4
Vals Finance IndexVals AI–46.5%
Vals Science IndexVals AI–22.0%
Vals Legal IndexVals AI–61.8%
AA Cyber IndexArtificial Analysis–33.7
Metadata
AA Openness IndexArtificial Analysis83.338.9

Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.

What changed

  • Nemotron 3.5 Super VL: Nemotron 3.5 Super VL cached input price moved from $0.060/M to $0.020/M (67% down).
  • Nemotron 3.5 Super VL: Nemotron 3.5 Super VL input price moved from $0.18/M to $0.20/M (11% up); Nemotron 3.5 Super VL cost per task moved from $0.52/M to $0.56/M (7% up).
  • Nemotron 3.5 Super VL: BenchLeader Index 48.9 → 45.8.
  • Nemotron 3.5 Super VL: BenchLeader Index 45.7 → 48.9.
  • Nemotron 3.5 Super VL: 1 new result (HiL-Bench); 1 score revision.
  • Beam 1.0: BenchLeader Index 45.7 → 46.2.
  • Beam 1.0: 1 new result (SciCode).
  • Nemotron 3.5 Super VL: Nemotron 3.5 Super VL input price moved from $0.20/M to $0.18/M (10% down).
  • Beam 1.0: 1 score revision.
  • Beam 1.0: BenchLeader Index 45.2 → 45.7.

Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs

Nemotron 3.5 Super VL vs Beam 1.0: questions

Is Nemotron 3.5 Super VL better than Beam 1.0?
Nemotron 3.5 Super VL and Beam 1.0 are neck and neck on quality (45.8 vs 45.9, within the ±9.7 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 16 Nov 2026, so check the category scores for your use.
Nemotron 3.5 Super VL vs Beam 1.0: which is cheaper, and will that change?
Nemotron 3.5 Super VL is cheaper today: $0.275 against $0.600 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Nemotron 3.5 Super VL or Beam 1.0 better for coding?
Beam 1.0 scores higher in coding (37 vs 34 on the category index, where 50 is average). On FrontierCode 2.0 (AA), the top shared coding benchmark, it is 6.1% for Nemotron 3.5 Super VL against 5.5% for Beam 1.0.
Which is faster, Nemotron 3.5 Super VL or Beam 1.0?
Beam 1.0 streams faster (288 vs 271 output tokens per second, Artificial Analysis 7-day median).
Which is better for long documents?
Beam 1.0 scores higher on long-context benchmarks (64 vs 59). Both accept 1,048,576 tokens of context.
More questions (3)
Which has lower latency?
Beam 1.0 answers first: 5.4 s against 13.7 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
Do independent benchmarks agree on Nemotron 3.5 Super VL vs Beam 1.0?
Across the 21 benchmarks both have, Nemotron 3.5 Super VL leads 5, Beam 1.0 leads 16. Every overall index that rates both puts Beam 1.0 ahead.
When was Nemotron 3.5 Super VL released?
Nemotron 3.5 Super VL was released on 15 Oct 2026.