Agentic
Directional only- Nemotron 3 Ultra
- 51.3
- Qwen3.6-35B-A3B
- 51.5
- Weighted basis
- 2 vs 1 rows
- Reading
- Directional only
Nemotron 3 Ultra vs Qwen3.6-35B-A3B
Updated August 30, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Qwen3.6-35B-A3B has the higher public score estimate, 51.55 versus 46.32, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
8 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Long documents
Prompts that approach the documented context limit
Nemotron 3 Ultra
Nemotron 3 Ultra has the larger documented context window.
Confidence: documented
Show secondary and unsupported calls
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
11 Nemotron 3 Ultra only
8 Shared
33 Qwen3.6-35B-A3B only
- Shared results
- 8
- Nemotron 3 Ultra only
- 11
- Qwen3.6-35B-A3B only
- 33
- Like-for-like categories
- 0 / 8
3 categories use different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.
Coding
Directional only- Nemotron 3 Ultra
- 58.3
- Qwen3.6-35B-A3B
- 73.8
- Weighted basis
- 2 vs 3 rows
- Reading
- Directional only
Knowledge
Directional only- Nemotron 3 Ultra
- 53.8
- Qwen3.6-35B-A3B
- 51.4
- Weighted basis
- 3 vs 4 rows
- Reading
- Directional only
Reasoning
Not comparable- Nemotron 3 Ultra
- 61.9
- Qwen3.6-35B-A3B
- Not measured
- Weighted basis
- 1 vs 0 rows
- Reading
- Not comparable
Math
Not comparable- Nemotron 3 Ultra
- Not measured
- Qwen3.6-35B-A3B
- 88.2
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Multilingual
Not comparable- Nemotron 3 Ultra
- 83.0
- Qwen3.6-35B-A3B
- Not measured
- Weighted basis
- 1 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- Nemotron 3 Ultra
- Not measured
- Qwen3.6-35B-A3B
- 76.3
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Instruction following
Not comparable- Nemotron 3 Ultra
- 81.7
- Qwen3.6-35B-A3B
- Not measured
- Weighted basis
- 1 vs 0 rows
- Reading
- Not comparable
| Category | Nemotron 3 Ultra | Qwen3.6-35B-A3B | Weighted basis | Reading |
|---|---|---|---|---|
| Agentic | 51.3 | 51.5 | Directional only2 vs 1 rows | Directional only |
| Coding | 58.3 | 73.8 | Directional only2 vs 3 rows | Directional only |
| Knowledge | 53.8 | 51.4 | Directional only3 vs 4 rows | Directional only |
| Reasoning | 61.9 | Not measured | Not comparable1 vs 0 rows | Not comparable |
| Math | Not measured | 88.2 | Not comparable0 vs 2 rows | Not comparable |
| Multilingual | 83.0 | Not measured | Not comparable1 vs 0 rows | Not comparable |
| Multimodal | Not measured | 76.3 | Not comparable0 vs 2 rows | Not comparable |
| Instruction following | 81.7 | Not measured | Not comparable1 vs 0 rows | Not comparable |
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
HLE
Knowledge
Nemotron 3 Ultra: 26.7%Qwen3.6-35B-A3B: 21.4%Normalized gap 5.3Nemotron 3 Ultra source Qwen3.6-35B-A3B sourceTerminal-Bench 2.0
Agentic
Nemotron 3 Ultra: 56.4%Qwen3.6-35B-A3B: 51.5%Normalized gap 4.9Nemotron 3 Ultra source Qwen3.6-35B-A3B sourceMMLU-Pro
Knowledge
Nemotron 3 Ultra: 86.8%Qwen3.6-35B-A3B: 85.2%Normalized gap 1.6Nemotron 3 Ultra source Qwen3.6-35B-A3B sourceSWE-bench Verified
Coding
Nemotron 3 Ultra: 71.9%Qwen3.6-35B-A3B: 73.4%Normalized gap 1.5Nemotron 3 Ultra source Qwen3.6-35B-A3B sourceGPQA
Knowledge
Nemotron 3 Ultra: 87%Qwen3.6-35B-A3B: 86%Normalized gap 1.0Nemotron 3 Ultra source Qwen3.6-35B-A3B source
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- Nemotron 3 Ultra
- Self-hosted; infrastructure cost varies
- Fits in one request
- Qwen3.6-35B-A3B
- API rate not published
- Fits in one request
Nemotron 3 Ultra has no comparable published API token rate. Qwen3.6-35B-A3B has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- Nemotron 3 Ultra
- Self-hosted; infrastructure cost varies
- Fits in one request
- Qwen3.6-35B-A3B
- API rate not published
- Fits in one request
Nemotron 3 Ultra has no comparable published API token rate. Qwen3.6-35B-A3B has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Nemotron 3 Ultra
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
- Qwen3.6-35B-A3B
- API rate not published
- Fits in one request
- Cached-input rate unavailable
Nemotron 3 Ultra has no comparable published API token rate. Qwen3.6-35B-A3B has no comparable published API token rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationNemotron 3 UltraQwen3.6-35B-A3B
Context window
Maximum documented context; output-token limits may be lower.
Nemotron 3 Ultra
1M
Qwen3.6-35B-A3B
262K
API model ID
Nemotron 3 Ultra
Not sourced
Qwen3.6-35B-A3B
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Nemotron 3 Ultra
No comparable hosted API rate
Qwen3.6-35B-A3B
No comparable hosted API rate
Documented inputs
Nemotron 3 Ultra
Not sourced
Qwen3.6-35B-A3B
Not sourced
Documented outputs
Nemotron 3 Ultra
Not sourced
Qwen3.6-35B-A3B
Not sourced
Provider availability
Nemotron 3 Ultra
Not sourced
Qwen3.6-35B-A3B
Not sourced
Reasoning profile
Nemotron 3 Ultra
Reasoning
Qwen3.6-35B-A3B
Reasoning
Weight access
Nemotron 3 Ultra
Open Weight
Qwen3.6-35B-A3B
Open Weight
License
Nemotron 3 Ultra
Open Weight
Qwen3.6-35B-A3B
Open Weight
Release date
Nemotron 3 Ultra
2026-06-04
Qwen3.6-35B-A3B
2026-04-15
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- Qwen3.6-35B-A3B has the higher public score estimate, 51.55 versus 46.32, but the 90% score intervals overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Nemotron 3 Ultra has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence52 rows
Agentic
Terminal-Bench 2.0
Nemotron 3 Ultra56.4%SourceQwen3.6-35B-A3B51.5%SourceNemotron 3 Ultra leads this result
PinchBench
Nemotron 3 Ultra90.0%SourceQwen3.6-35B-A3B—Not directly comparable
BrowseComp
Nemotron 3 Ultra44.4%SourceQwen3.6-35B-A3B—Not directly comparable
τ³-bench results
Nemotron 3 Ultra70.9%SourceQwen3.6-35B-A3B67.2%SourceNemotron 3 Ultra leads this result
HLE w/ tools
Nemotron 3 Ultra37.4%SourceQwen3.6-35B-A3B—Not directly comparable
Claw-Eval
Nemotron 3 Ultra—Qwen3.6-35B-A3B68.7%SourceNot directly comparable
QwenClawBench
Nemotron 3 Ultra—Qwen3.6-35B-A3B52.6%SourceNot directly comparable
QwenWebBench
Nemotron 3 Ultra—Qwen3.6-35B-A3B1397SourceNot directly comparable
VITA-Bench
Nemotron 3 Ultra—Qwen3.6-35B-A3B35.6%SourceNot directly comparable
DeepPlanning
Nemotron 3 Ultra—Qwen3.6-35B-A3B25.9%SourceNot directly comparable
Toolathlon
Nemotron 3 Ultra—Qwen3.6-35B-A3B26.9%SourceNot directly comparable
MCP Atlas
Nemotron 3 Ultra—Qwen3.6-35B-A3B62.8%SourceNot directly comparable
WideResearch
Nemotron 3 Ultra—Qwen3.6-35B-A3B60.1%SourceNot directly comparable
Gert Labs
Nemotron 3 Ultra—Qwen3.6-35B-A3B42.65%SourceNot directly comparable
Coding
SWE-bench Verified
Nemotron 3 Ultra71.9%SourceQwen3.6-35B-A3B73.4%SourceQwen3.6-35B-A3B leads this result
SWE Multilingual
Nemotron 3 Ultra67.7%SourceQwen3.6-35B-A3B67.2%SourceNemotron 3 Ultra leads this result
LiveCodeBench v6
Nemotron 3 Ultra89.0%SourceQwen3.6-35B-A3B—Not directly comparable
SciCode
Nemotron 3 Ultra44.6%SourceQwen3.6-35B-A3B—Not directly comparable
Terminal-Bench 2.0
Nemotron 3 Ultra56.4%SourceQwen3.6-35B-A3B51.5%SourceNemotron 3 Ultra leads this result
SWE-bench Pro
Nemotron 3 Ultra—Qwen3.6-35B-A3B49.5%SourceNot directly comparable
LiveCodeBench
Nemotron 3 Ultra—Qwen3.6-35B-A3B80.4%SourceNot directly comparable
NL2Repo
Nemotron 3 Ultra—Qwen3.6-35B-A3B29.4%SourceNot directly comparable
Reasoning
Knowledge
GPQA
Nemotron 3 Ultra87%SourceQwen3.6-35B-A3B86%SourceNemotron 3 Ultra leads this result
GPQA-D
Nemotron 3 Ultra87.0%SourceQwen3.6-35B-A3B—Not directly comparable
HLE
Nemotron 3 Ultra26.7%SourceQwen3.6-35B-A3B21.4%SourceNemotron 3 Ultra leads this result
HLE w/o tools
Nemotron 3 Ultra26.7%SourceQwen3.6-35B-A3B—Not directly comparable
MMLU-Pro
Nemotron 3 Ultra86.8%SourceQwen3.6-35B-A3B85.2%SourceNemotron 3 Ultra leads this result
SuperGPQA
Nemotron 3 Ultra—Qwen3.6-35B-A3B64.7%SourceNot directly comparable
C-Eval
Nemotron 3 Ultra—Qwen3.6-35B-A3B90%SourceNot directly comparable
Math
HMMT Feb 2025
Nemotron 3 Ultra—Qwen3.6-35B-A3B90.7%SourceNot directly comparable
HMMT Nov 2025
Nemotron 3 Ultra—Qwen3.6-35B-A3B89.1%SourceNot directly comparable
HMMT Feb 2026
Nemotron 3 Ultra—Qwen3.6-35B-A3B83.6%SourceNot directly comparable
MMAnswerBench
Nemotron 3 Ultra—Qwen3.6-35B-A3B78.9%SourceNot directly comparable
AIME26
Nemotron 3 Ultra—Qwen3.6-35B-A3B92.7%SourceNot directly comparable
Multilingual
MMLU-ProX
Nemotron 3 Ultra83%SourceQwen3.6-35B-A3B—Not directly comparable
Multimodal
MMMU
Nemotron 3 Ultra—Qwen3.6-35B-A3B81.7%SourceNot directly comparable
MMMU-Pro
Nemotron 3 Ultra—Qwen3.6-35B-A3B75.3%SourceNot directly comparable
RealWorldQA
Nemotron 3 Ultra—Qwen3.6-35B-A3B85.3%SourceNot directly comparable
OmniDocBench 1.5
Nemotron 3 Ultra—Qwen3.6-35B-A3B89.9%SourceNot directly comparable
CharXiv
Nemotron 3 Ultra—Qwen3.6-35B-A3B78%SourceNot directly comparable
SimpleVQA
Nemotron 3 Ultra—Qwen3.6-35B-A3B58.9%SourceNot directly comparable
CC-OCR
Nemotron 3 Ultra—Qwen3.6-35B-A3B81.9%SourceNot directly comparable
AI2D_TEST
Nemotron 3 Ultra—Qwen3.6-35B-A3B92.7%SourceNot directly comparable
RefCOCO (avg)
Nemotron 3 Ultra—Qwen3.6-35B-A3B92.0%SourceNot directly comparable
ODINW13
Nemotron 3 Ultra—Qwen3.6-35B-A3B50.8%SourceNot directly comparable
Video-MME (with subtitle)
Nemotron 3 Ultra—Qwen3.6-35B-A3B86.6%SourceNot directly comparable
Video-MME (w/o subtitle)
Nemotron 3 Ultra—Qwen3.6-35B-A3B82.5%SourceNot directly comparable
VideoMMMU
Nemotron 3 Ultra—Qwen3.6-35B-A3B83.7%SourceNot directly comparable
MLVU (M-Avg)
Nemotron 3 Ultra—Qwen3.6-35B-A3B86.2%SourceNot directly comparable
Instruction following
IFBench
Nemotron 3 Ultra81.7%SourceQwen3.6-35B-A3B—Not directly comparable
Frequently asked questions
Which is better, Nemotron 3 Ultra or Qwen3.6-35B-A3B?
Qwen3.6-35B-A3B has the higher public score estimate, 51.55 versus 46.32, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, Nemotron 3 Ultra or Qwen3.6-35B-A3B?
The current coding averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which is better for agentic tasks, Nemotron 3 Ultra or Qwen3.6-35B-A3B?
The current agentic tasks averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which costs less, Nemotron 3 Ultra or Qwen3.6-35B-A3B?
Both models do not have comparable published API token rates, so this page does not name a universal price winner.
Which has the larger context window, Nemotron 3 Ultra or Qwen3.6-35B-A3B?
Nemotron 3 Ultra has the larger documented context window: 1M, compared with 262K.
Related comparisons
Last updated August 30, 2026
Watch Nemotron 3 Ultra vs Qwen3.6-35B-A3B
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.