Claude Opus 4.6 vs Edge0-145B-NVFP4 vs NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB
New ComparisonScores are fetched live from the current database on every visit — nothing is cached.
Winner
Claude Opus 4.6 scores higher overall across category strength profiles (95 vs 40 for NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB); only 1 benchmarks are directly comparable.
Only 1 of 45 benchmarks are scored by every selected model — the verdict leans on category strength profiles and is indicative.
Benchmark Scores Comparison
Claude Opus 4.6
Score: 15.0
Edge0-145B-NVFP4
145.0B params
NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB
74.0B params
Coverage matrix
3 models × 45 benchmarks — hover a benchmark name for its category
| Benchmark | NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB NVIDIA | Edge0-145B-NVFP4 Edge0 | Claude Opus 4.6 Anthropic |
|---|---|---|---|
| OmniScience Accuracy | 16.9 | — | — |
| GPQA Diamond | 86.43 | 90.4 | 91.3 |
| LiveCodeBench v6 | — | — | 88.8 |
| SWE-bench Pro | — | — | 53.4 |
| AA-LCR | 64.06 | 79.3 | — |
| MathVision | — | — | 71.2 |
| AIME 26 | — | — | 96.7 |
| AgentWorldBench Terminal | — | — | 57.51 |
| Toolathlon Verified | — | — | 47.2 |
| OJBench | — | — | 60.3 |
| AgentWorldBench Overall | — | — | 57.8 |
| OmniScience Non-Hallucination | 61.27 | — | — |
| DeepSearch QA | — | — | 91.3 |
| AgentWorldBench Web | — | — | 51.42 |
| MMMU-Pro | — | — | 73.9 |
| BabyVision | — | — | 14.8 |
| Claw-Eval Avg | — | — | 82.4 |
| Apex-Agents | — | — | 33 |
| PinchBench | 87.82 | — | — |
| OSWorld-Verified | — | — | 72.7 |
| AgentWorldBench Search | — | — | 29.3 |
| HLE (with tools) | — | 35.4 | 53 |
| CharXiv (RQ) | — | — | 69.1 |
| BrowseComp | 66.75 | — | 83.7 |
| MCPMark | — | — | 56.7 |
| SWE-bench Multilingual | — | — | 77.8 |
| IFBench | — | 80.6 | — |
| SWE-bench Verified | — | — | 80.8 |
| IMOAnswerBench | — | — | 75.3 |
| Humanity's Last Exam | 25.3 | — | 40 |
| Terminal-Bench 2.0 | — | — | 65.4 |
| Terminal-Bench 2.1 (Terminus-2) | — | 86.5 | — |
| HMMT Feb 26 | — | — | 96.2 |
| AgentWorldBench OS | — | — | 70.2 |
| AgentWorldBench MCP | — | — | 69.9 |
| Terminal Bench 2.1 | 50.98 | — | — |
| AgentWorldBench SWE | — | — | 64.55 |
| SciCode (subtask) | — | — | 51.9 |
| Claw-Eval Pass^3 | — | — | 70.4 |
| MCP-Atlas | 52.2 | — | — |
| V-Star | — | — | 86.4 |
| AgentWorldBench Android | — | — | 61.74 |
| DeepSWE 1.1 | — | 64.6 | — |
| SciCode | 41.02 | — | — |
| GDPVal-AA v2 | 28.47 | — | — |
| Covered | 11/45 | 6/45 | 34/45 |
Scorecard
Overall 0–100 score — category strengths (50%), shared-benchmark wins (30%), coverage (20%)
| Model | Overall score | Shared benchmark wins | Coverage | Strongest at | Weakest at |
|---|---|---|---|---|---|
| Claude Opus 4.6 Winner | 95.1 | 1 of 1 | 34 of 45 (76.0%) | STEM & Reasoning | General Agents |
| NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB | 40.2 | 0 of 1 | 11 of 45 (24.0%) | Long Context | STEM & Reasoning |
| Edge0-145B-NVFP4 | 39.6 | 0 of 1 | 6 of 45 (13.0%) | Long Context | STEM & Reasoning |
Category strengths
Where each model stands, per benchmark category — score relative to the best model, adjusted for benchmark coverage. This also ranks models when they share few benchmarks.
| Category | Claude Opus 4.6 | NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB | Edge0-145B-NVFP4 |
|---|---|---|---|
| General Agents 17 benchmarks | 100 | 79 | — |
| STEM & Reasoning 9 benchmarks | 100 | 64.4 | 65.1 |
| Long Context 1 benchmark | — | 80.8 | 100 |
Fully comparable benchmarks
Scored by all 3 selected models (1)
| Benchmark | Category | Claude Opus 4.6 | Edge0-145B-NVFP4 | NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB |
|---|---|---|---|---|
| GPQA Diamond | stem_reasoning | 91.3 | 90.4 | 86.43 |
Partial coverage benchmarks
Scored by at least one but not all selected models (44) — missing scores show as —
| Benchmark | Category | Claude Opus 4.6 | Edge0-145B-NVFP4 | NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB |
|---|---|---|---|---|
| OmniScience Accuracy 1/3 models | knowledge | — | — | 16.9 |
| LiveCodeBench v6 1/3 models | stem_reasoning | 88.8 | — | — |
| SWE-bench Pro 1/3 models | coding_agent | 53.4 | — | — |
| AA-LCR 2/3 models | long_context | — | 79.3 | 64.06 |
| MathVision 1/3 models | vision_language | 71.2 | — | — |
| AIME 26 1/3 models | stem_reasoning | 96.7 | — | — |
| AgentWorldBench Terminal 1/3 models | general_agent | 57.51 | — | — |
| Toolathlon Verified 1/3 models | general_agent | 47.2 | — | — |
| OJBench 1/3 models | stem_reasoning | 60.3 | — | — |
| AgentWorldBench Overall 1/3 models | general_agent | 57.8 | — | — |
| OmniScience Non-Hallucination 1/3 models | knowledge | — | — | 61.27 |
| DeepSearch QA 1/3 models | general_agent | 91.3 | — | — |
| AgentWorldBench Web 1/3 models | general_agent | 51.42 | — | — |
| MMMU-Pro 1/3 models | vision_language | 73.9 | — | — |
| BabyVision 1/3 models | vision_language | 14.8 | — | — |
| Claw-Eval Avg 1/3 models | coding_agent | 82.4 | — | — |
| Apex-Agents 1/3 models | general_agent | 33 | — | — |
| PinchBench 1/3 models | general_agent | — | — | 87.82 |
| OSWorld-Verified 1/3 models | general_agent | 72.7 | — | — |
| AgentWorldBench Search 1/3 models | general_agent | 29.3 | — | — |
| HLE (with tools) 2/3 models | stem_reasoning | 53 | 35.4 | — |
| CharXiv (RQ) 1/3 models | document_understanding | 69.1 | — | — |
| BrowseComp 2/3 models | general_agent | 83.7 | — | 66.75 |
| MCPMark 1/3 models | general_agent | 56.7 | — | — |
| SWE-bench Multilingual 1/3 models | coding_agent | 77.8 | — | — |
| IFBench 1/3 models | instruction_following | — | 80.6 | — |
| SWE-bench Verified 1/3 models | coding_agent | 80.8 | — | — |
| IMOAnswerBench 1/3 models | stem_reasoning | 75.3 | — | — |
| Humanity's Last Exam 2/3 models | stem_reasoning | 40 | — | 25.3 |
| Terminal-Bench 2.0 1/3 models | coding_agent | 65.4 | — | — |
| Terminal-Bench 2.1 (Terminus-2) 1/3 models | coding_agent | — | 86.5 | — |
| HMMT Feb 26 1/3 models | stem_reasoning | 96.2 | — | — |
| AgentWorldBench OS 1/3 models | general_agent | 70.2 | — | — |
| AgentWorldBench MCP 1/3 models | general_agent | 69.9 | — | — |
| Terminal Bench 2.1 1/3 models | coding_agent | — | — | 50.98 |
| AgentWorldBench SWE 1/3 models | general_agent | 64.55 | — | — |
| SciCode (subtask) 1/3 models | stem_reasoning | 51.9 | — | — |
| Claw-Eval Pass^3 1/3 models | coding_agent | 70.4 | — | — |
| MCP-Atlas 1/3 models | general_agent | — | — | 52.2 |
| V-Star 1/3 models | vision_language | 86.4 | — | — |
| AgentWorldBench Android 1/3 models | general_agent | 61.74 | — | — |
| DeepSWE 1.1 1/3 models | coding_agent | — | 64.6 | — |
| SciCode 1/3 models | — | — | — | 41.02 |
| GDPVal-AA v2 1/3 models | general_agent | — | — | 28.47 |
Comparison FAQ
Which model scores highest in this comparison?
Claude Opus 4.6 scores higher overall across category strength profiles (95 vs 40 for NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB); only 1 benchmarks are directly comparable.
Which model is strongest at STEM & Reasoning?
Claude Opus 4.6 leads the STEM & Reasoning category (strength score 100/100 across 3 comparable benchmarks).
How reliable is this comparison verdict?
Only 1 of 45 benchmarks are scored by every selected model — the verdict leans on category strength profiles and is indicative.