Model comparison
GPT-5.4 vs Nemotron 3 Ultra
Head-to-head evidence from 23 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Verified leaderboard positions: GPT-5.4 #10; Nemotron 3 Ultra #24
BenchAlign evidence: GPT-5.4 supported; Nemotron 3 Ultra not scored. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. GPT-5.4 and Nemotron 3 Ultra share 23 comparable benchmark results. 3 of 8 categories are comparable. 31 results are unique to GPT-5.4; 16 to Nemotron 3 Ultra.
Updated July 16, 2026- Shared results
- 23
- GPT-5.4 only
- 31
- Nemotron 3 Ultra only
- 16
- Comparable categories
- 3 / 8
Pick GPT-5.4 if you want the stronger benchmark profile. Nemotron 3 Ultra only becomes the better choice if coding is the priority or you want the cheaper token bill.
Confidence note. This is a partial-evidence comparison with 23 shared benchmark results across 6 evidence categories; 3 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
GPT-5.4 is clearly ahead on the provisional aggregate, 86 to 60. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
GPT-5.4's sharpest advantage is in agentic, where it averages 77.2 against 51.3. The single biggest benchmark swing on the page is BrowseComp, 82.7% to 44.4%. Nemotron 3 Ultra does hit back in coding, so the answer changes if that is the part of the workload you care about most.
GPT-5.4 is also the more expensive model on tokens at $2.50 input / $15.00 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Nemotron 3 Ultra. That is roughly Infinityx on output cost alone. GPT-5.4 gives you the larger context window at 1.05M, compared with 1M for Nemotron 3 Ultra.
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | GPT-5.4 | Δ | Nemotron 3 Ultra |
|---|---|---|---|
| Agentic | GPT-5.477.2 | Margin← 25.9 | Nemotron 3 Ultra51.3 |
| Knowledge | GPT-5.457.6 | Margin← 3.4 | Nemotron 3 Ultra54.2 |
| Coding | GPT-5.457.7 | Margin→ 0.6 | Nemotron 3 Ultra58.3 |
| Reasoning | GPT-5.4Not measured | MarginNo overlap | Nemotron 3 Ultra61.9 |
| Math | GPT-5.442.5 | MarginNo overlap | Nemotron 3 UltraNot measured |
| Multilingual | GPT-5.4Not measured | MarginNo overlap | Nemotron 3 Ultra83.0 |
| Multimodal | GPT-5.473.2 | MarginNo overlap | Nemotron 3 UltraNot measured |
| Inst. Following | GPT-5.4Not measured | MarginNo overlap | Nemotron 3 Ultra81.7 |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
BrowseComp
AgenticA 82.7%B 44.4%Winner: GPT-5.4Δ 38.3BrowseComp: GPT-5.4 scored 82.7%; Nemotron 3 Ultra scored 44.4%. GPT-5.4 wins this benchmark. - Source ↗
HLE
KnowledgeA 52.1%B 26.7%Winner: GPT-5.4Δ 25.4HLE: GPT-5.4 scored 52.1%; Nemotron 3 Ultra scored 26.7%. GPT-5.4 wins this benchmark. - Source ↗
Terminal-Bench 2.0
AgenticA 75.1%B 56.4%Winner: GPT-5.4Δ 18.7Terminal-Bench 2.0: GPT-5.4 scored 75.1%; Nemotron 3 Ultra scored 56.4%. GPT-5.4 wins this benchmark. - Source ↗
GPQA
KnowledgeA 92.8%B 87%Winner: GPT-5.4Δ 5.8GPQA: GPT-5.4 scored 92.8%; Nemotron 3 Ultra scored 87%. GPT-5.4 wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | GPT-5.4 | Nemotron 3 Ultra | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | GPT-5.4$2.5 input / $15 output | Nemotron 3 Ultra$0 input / $0 output | Nemotron 3 Ultra has the lower combined listed price. |
| Generation speedtokens per second | GPT-5.474 tok/s | Nemotron 3 UltraNot available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | GPT-5.4151.79 s | Nemotron 3 UltraNot available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | GPT-5.41.05M | Nemotron 3 Ultra1M | GPT-5.4 lists the larger context window. |
Benchmark Deep Dive
AgenticGPT-5.4 wins23 benchmarks
| Benchmark | GPT-5.4 | Nemotron 3 Ultra | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 75.1% | 56.4% | GPT-5.4 leads |
| CyberGymSource | 79.0% | — | Not comparable |
| BrowseCompSource | 82.7% | 44.4% | GPT-5.4 leads |
| OSWorld-VerifiedSource | 75% | — | Not comparable |
| MCP AtlasSource | 70.6% | — | Not comparable |
| ToolathlonSource | 54.6% | — | Not comparable |
| τ²-bench resultsSource | 87.1% | 83.3% | GPT-5.4 leads |
| Claw-EvalSource | 60.3% | — | Not comparable |
| DeepSearchQASource | 73.6% | — | Not comparable |
| AA Agentic IndexSource | 41.1% | 27.4% | GPT-5.4 leads |
| APEX-Agents-AASource | 33.3% | — | Not comparable |
| GDPval-AASource | 44.7% | 33.2% | GPT-5.4 leads |
| GDPval-AASource | 1395 | 1164 | GPT-5.4 leads |
| Gert LabsSource | 64.89% | — | Not comparable |
| ResearchClawBenchSource | 15.3% | — | Not comparable |
| JobBenchSource | 38.9% | — | Not comparable |
| ExploitGymSource | 6.0% | — | Not comparable |
| PinchBenchSource | — | 90.0% | Not comparable |
| τ³-bench resultsSource | — | 70.9% | Not comparable |
| HLE w/ toolsSource | — | 37.4% | Not comparable |
| AA BriefcaseSource | — | 870 | Not comparable |
| AA EnterpriseOps-GymSource | — | 28.9% | Not comparable |
| AA Harvey LABSource | — | 3.3% | Not comparable |
CodingNemotron 3 Ultra wins12 benchmarks
| Benchmark | GPT-5.4 | Nemotron 3 Ultra | Result |
|---|---|---|---|
| LiveCodeBench ProSource | 87.5% | — | Not comparable |
| SWE-bench ProSource | 57.7% | — | Not comparable |
| React Native EvalsSource | 85.3% | — | Not comparable |
| Vibe Code BenchSource | 67.42% | — | Not comparable |
| AA Coding IndexSource | 71.0% | 49.3% | GPT-5.4 leads |
| Terminal-Bench HardSource | 57.6% | 36.4% | GPT-5.4 leads |
| AA-SciCodeSource | 56.6% | 39.9% | GPT-5.4 leads |
| SWE-bench VerifiedSource | — | 71.9% | Not comparable |
| SWE MultilingualSource | — | 67.7% | Not comparable |
| LiveCodeBench v6Source | — | 89.0% | Not comparable |
| SciCodeSource | — | 44.6% | Not comparable |
| Terminal-Bench 2.0Source | — | 56.4% | Not comparable |
Reasoning3 benchmarks
KnowledgeGPT-5.4 wins15 benchmarks
| Benchmark | GPT-5.4 | Nemotron 3 Ultra | Result |
|---|---|---|---|
| GPQASource | 92.8% | 87% | GPT-5.4 leads |
| HLESource | 52.1% | 26.7% | GPT-5.4 leads |
| HLE w/o toolsSource | 39.8% | 26.7% | GPT-5.4 leads |
| GPQA-DSource | 92.8% | 87.0% | GPT-5.4 leads |
| HealthBench HardSource | 40.1% | — | Not comparable |
| MedXpertQA (Text)Source | 59.6% | — | Not comparable |
| Artificial Analysis Intelligence IndexSource | 51.4% | 37.8% | GPT-5.4 leads |
| AA-GPQA DiamondSource | 92.0% | 86.7% | GPT-5.4 leads |
| AA-HLESource | 41.6% | 26.6% | GPT-5.4 leads |
| AA-Omniscience IndexSource | 5.7% | -0.8% | GPT-5.4 leads |
| AA-Omniscience AccuracySource | 50.0% | 21.6% | GPT-5.4 leads |
| AA-Omniscience Hallucination RateSource | 88.6% | 28.5% | Nemotron 3 Ultra leads |
| HealthBench ProfessionalSource | 48.1% | — | Not comparable |
| MMLU-ProSource | — | 86.8% | Not comparable |
| AA Openness IndexSource | — | 83.3% | Not comparable |
Math2 benchmarks
Multilingual1 benchmarks
| Benchmark | GPT-5.4 | Nemotron 3 Ultra | Result |
|---|---|---|---|
| MMLU-ProXSource | — | 83% | Not comparable |
Multimodal12 benchmarks
| Benchmark | GPT-5.4 | Nemotron 3 Ultra | Result |
|---|---|---|---|
| MMMU-ProSource | 81.2% | — | Not comparable |
| OfficeQA ProSource | 53.2% | — | Not comparable |
| MMMU-Pro w/ PythonSource | 82.1% | — | Not comparable |
| CharXivSource | 82.8% | — | Not comparable |
| ERQASource | 65.4% | — | Not comparable |
| SimpleVQASource | 61.1% | — | Not comparable |
| ScreenSpot ProSource | 85.4% | — | Not comparable |
| ZeroBenchSource | 41.0% | — | Not comparable |
| MedXpertQA (MM)Source | 77.1% | — | Not comparable |
| GDPval-AASource | 1672 | — | Not comparable |
| AA-MMMU-ProSource | 78.4% | — | Not comparable |
| Design Arena WebsiteSource | 1254 | 1132 | GPT-5.4 leads |
Frequently Asked Questions (4)
Which is better, GPT-5.4 or Nemotron 3 Ultra?
GPT-5.4 is ahead on BenchLM's provisional leaderboard, 86 to 60. The biggest single separator in this matchup is BrowseComp, where the scores are 82.7% and 44.4%.
Which is better for knowledge tasks, GPT-5.4 or Nemotron 3 Ultra?
GPT-5.4 has the edge for knowledge tasks in this comparison, averaging 57.6 versus 54.2. Inside this category, AA-Omniscience Hallucination Rate is the benchmark that creates the most daylight between them.
Which is better for coding, GPT-5.4 or Nemotron 3 Ultra?
Nemotron 3 Ultra has the edge for coding in this comparison, averaging 58.3 versus 57.7. Inside this category, AA Coding Index is the benchmark that creates the most daylight between them.
Which is better for agentic tasks, GPT-5.4 or Nemotron 3 Ultra?
GPT-5.4 has the edge for agentic tasks in this comparison, averaging 77.2 versus 51.3. Inside this category, GDPval-AA is the benchmark that creates the most daylight between them.
Related Comparisons
Explore More
The AI models change fast. We track them for you.
A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.
Free. No spam. Unsubscribe anytime.