Compare / head-to-head
GPT-5.5vsNVIDIA Nemotron 3 Super 120B-A12B
GPT-5.5 leads 1 of 1 shared benchmark. Both offer a 1M-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
1 – 0
GPT-5.5 leads
Cheaper per token
—
list price, input + output
Larger context
Tie
both 1M tokens
Provider uptime (30d)
94.4637% · 100%
OpenAI · NVIDIA
GPT-5.5
OpenAI · released 2026-04-23
textvision
NVIDIA Nemotron 3 Super 120B-A12B
NVIDIA · released 2026-03-11
text
- Context
- 1M
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 10 · 8 core
Quality
Benchmark matrix
1 shared · 20 only GPT-5.5 · 9 only NVIDIA Nemotron 3 Super 120B-A12B| Benchmark | GPT-5.5 | NVIDIA Nemotron 3 Super 120B-A12B | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 1 | ||||
| LMArena Elo | 1465.6 ↗ | 1377.9 ↗ | +87.7 | GPT-5.5 |
| Only GPT-5.5 reports · 20 | ||||
| GPQA Diamond | 90.7% ↗epoch | not reported | — | — |
| AIME 2026 | 100% ↗matharena ⚠ | not reported | — | — |
| ARC-AGI-1 (Verified) | 95% ↗ | not reported | — | — |
| ARC-AGI-2 (Verified) | 85% ↗ | not reported | — | — |
| BrowseComp | 84.4% ↗ | not reported | — | — |
| FrontierMath Tier 1-3 | 51.7% ↗ | not reported | — | — |
| FrontierMath Tier 4 | 35.4% ↗ | not reported | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | 72.5% ↗epoch | not reported | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 85.3% ↗epoch | not reported | — | — |
| GDPval (wins or ties) | 84.9% ↗ | not reported | — | — |
| MCP Atlas | 75.3% ↗ | not reported | — | — |
| MMMU-Pro (no tools) | 81.2% ↗ | not reported | — | — |
| MMMU-Pro (with tools) | 83.2% ↗ | not reported | — | — |
| OSWorld-Verified | 78.7% ↗ | not reported | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 84.4% ↗epoch | not reported | — | — |
| SimpleQA Verified | 63% ↗epoch | not reported | — | — |
| SWE-Bench Pro | 58.6% ↗ | not reported | — | — |
| Tau2-bench Telecom | 98% ↗ | not reported | — | — |
| Terminal-Bench 2.0 | 82.7% ↗ | not reported | — | — |
| Toolathlon | 55.6% ↗ | not reported | — | — |
| Only NVIDIA Nemotron 3 Super 120B-A12B reports · 9 | ||||
| SWE-bench Verified | not reported | 60.47% ↗ | — | — |
| MMLU-Pro | not reported | 83.73% ↗ | — | — |
| AIME 2025 | not reported | 90.21% ↗ | — | — |
| Humanity's Last Exam (no tools) | not reported | 18.26% ↗ | — | — |
| GPQA (no tools) | not reported | 79.23% ↗ | — | — |
| GPQA (with tools) | not reported | 82.7% ↗ | — | — |
| Humanity's Last Exam (with tools) | not reported | 22.82% ↗ | — | — |
| LiveCodeBench | not reported | 81.19% ↗ | — | — |
| Terminal-Bench Core 2.0 | not reported | 31% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is GPT-5.5 minus NVIDIA Nemotron 3 Super 120B-A12B in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | GPT-5.5 | NVIDIA Nemotron 3 Super 120B-A12B | Edge |
|---|---|---|---|
| Context window | 1M tokens | 1M tokens | Tie |
| Max output | 128K tokens | — | — |
| Input price / 1M | $5 ↗ | — | — |
| Output price / 1M | $30 ↗ | — | — |
| Cached input / 1M | $0.5 ↗ | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $35 | — | — |
| Modalities | text · vision | text | GPT-5.5 |
| Released | 2026-04-23 | 2026-03-11 | — |
| Cited benchmark scores | 21 | 10 | — |
Reliability
Provider status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
GPT-5.5 vs …
Models sharing the most benchmarksMore matchups
NVIDIA Nemotron 3 Super 120B-A12B vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run GPT-5.5 and NVIDIA Nemotron 3 Super 120B-A12B on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.