NVIDIAOpen weights: nvidia/NVIDIA-Nemotron-3.5-Super-VLReasoning modelNewReleased 15 Oct 2026Compare with another modelNemotron 3.5 Super VL
Nemotron 3.5 Super VL is a NVIDIA open-weights reasoning model, released 15 Oct 2026. It is not yet ranked: 4 independent results so far, and the index needs at least five across two categories. At $0.315 per million tokens blended it is cheaper than most ranked models. Output speed of 409 tokens per second puts it in the fastest quarter, with a first answer in 5.6 s. Last measured 15 Oct 2026.
- Blended price
- $0.315/M
- $0.230 in · $0.570 out
- Output speed
- 409 tok/s
- measured by Artificial Analysis
- First answer
- 5.61 s
- first token 0.76 s
- Context
- 1.0M
- Full answer
- 6.83 s
- median, reasoning included
- Cost per run
- $0.188
- one full Intelligence Index run
- Released
- 15 Oct 2026
- Reasoning58
- Knowledge60
- Long context59
- Composite59
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | not stated | Source |
|---|---|---|
| Humanity's Last Exam (AA)not in index | 30.7%#140 | Artificial Analysis |
| CritPt | 10.6%#113 | Artificial Analysis |
Coding
| Benchmark | not stated | Source |
|---|---|---|
| SciCode (AA)not in index | 47.8%#125 | Artificial Analysis |
| Terminal-Bench 4.0 (AA)not in index | 0.0%#147 | Artificial Analysis |
Agents & tools
| Benchmark | not stated | Source |
|---|---|---|
| GDPval-AA v2.1not in index | 33.1%#124 | Artificial Analysis |
| AutomationBenchnot in index | 14.8%#139 | Artificial Analysis |
| GDP.pdfnot in index | 8.0%#140 | Artificial Analysis |
| AA-Briefcase v1.1not in index | 999#49 | Artificial Analysis |
Knowledge
| Benchmark | not stated | Source |
|---|---|---|
| AA-Omniscience | -4.2#154 | Artificial Analysis |
| AA-Omniscience: accuracynot in index | 23.3%#251 | Artificial Analysis |
| AA-Omniscience: non-hallucinationnot in index | 64.2%#67 | Artificial Analysis |
Long context
| Benchmark | not stated | Source |
|---|---|---|
| AA-LCR | 69.0%#227 | Artificial Analysis |
Composite
| Benchmark | not stated | Source |
|---|---|---|
| AA Intelligence Index v4.3.2 | 26.5#148 | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| CoreWeave | 216 tok/s | 0.51 s | $0.150 | $0.500 | 1.0M | nvfp4 |
| Venice | 195 tok/s | 0.65 s | $0.250 | $1.25 | 262k | fp4 |
| DeepInfra | 192 tok/s | 0.76 s | $0.100 | $0.400 | 262k | fp4 |
| Together | 113 tok/s | 0.58 s | $0.150 | $0.500 | 262k | – |
| Crusoe | 5 tok/s | 24 s | $0.100 | $0.400 | 262k | bf16 |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.300 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0003 | $0.0003 | 6.3 s |
| Summarise a 30-page report | 12,000 / 600 | $0.0031 | $0.0037 | 7.1 s |
| Code edit | 6,000 / 1,500 | $0.0022 | $0.0025 | 9.3 s |
| Agentic coding session | 60,000 / 4,000 | $0.016 | $0.019 | 15.4 s |
| Structured extraction | 2,000 / 200 | $0.0006 | $0.0007 | 6.1 s |
More from NVIDIA
See also
Data as of 18 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Nemotron 3.5 Super VL: benchmarks, pricing, speed and rank”, https://www.benchleader.com/models/nemotron-3-5-super-vl, data as of 18 Oct 2026.