DeepSeek-V4-ProvsNemotron 3 Ultra
Across 9 shared benchmarks, DeepSeek-V4-Pro leads overall: DeepSeek-V4-Pro wins 5, Nemotron 3 Ultra wins 4, with 0 ties and an average score difference of -2.83.
DeepSeek-V4-Pro
DeepSeek-AI · 2026-08-13 · Reasoning model
Nemotron 3 Ultra
NVIDIA · 2026-06-04 · Reasoning model
DeepSeek-V4-Pro5 wins(56%)(44%)4 winsNemotron 3 Ultra
Benchmark scores
Grouped by capability, sorted by largest gap within each. 9 shared benchmarks.
Coding and Software Engineer
DeepSeek-V4-Pro 2/3| Benchmark | DeepSeek-V4-Pro | Nemotron 3 Ultra | Diff |
|---|---|---|---|
| LiveCodeBench | 56.8079 / 126Normal (No Tools) | 899 / 126Thinking (No Tools) | -32.20 |
| SWE-bench Verified | 73.6046 / 114Normal (With Tools) | 70.7058 / 114Thinking (With Tools) | +2.90 |
| SWE-bench Multilingual | 69.8019 / 25Normal (With Tools) | 67.7022 / 25Thinking (With Tools) | +2.10 |
General Knowledge
Nemotron 3 Ultra 2/3| Benchmark | DeepSeek-V4-Pro | Nemotron 3 Ultra | Diff |
|---|---|---|---|
| HLE | 7.70165 / 181Normal (No Tools) | 37.4073 / 181Thinking (With Tools) | -29.70 |
| LiveBench | 73.5823 / 115Normal (No Tools) | 51.7888 / 115Normal (No Tools) | +21.80 |
| MMLU Pro | 82.9049 / 133Normal (No Tools) | 86.8016 / 133Thinking (No Tools) | -3.90 |
AI Agent - Information Search
DeepSeek-V4-Pro 1/1| Benchmark | DeepSeek-V4-Pro | Nemotron 3 Ultra | Diff |
|---|---|---|---|
| BrowseComp | 83.4013 / 54极高强度思考（工具） | 44.4046 / 54Thinking (With Tools + Internet) | +39 |
AI Agent - Tool Usage
DeepSeek-V4-Pro 1/1| Benchmark | DeepSeek-V4-Pro | Nemotron 3 Ultra | Diff |
|---|---|---|---|
| Terminal-Bench 2.1 | 87.906 / 44极高强度思考（工具） | 56.4041 / 44Thinking (With Tools) | +31.50 |
Math and Reasoning
Nemotron 3 Ultra 1/1| Benchmark | DeepSeek-V4-Pro | Nemotron 3 Ultra | Diff |
|---|---|---|---|
| IMO-AnswerBench | 35.3023 / 23Normal (No Tools) | 92.301 / 23Thinking (With Tools) | -57 |
Specs
| Field | DeepSeek-V4-Pro | Nemotron 3 Ultra |
|---|---|---|
| Publisher | DeepSeek-AI | NVIDIA |
| Release date | 2026-08-13 | 2026-06-04 |
| Model type | Reasoning model | Reasoning model |
| Architecture | MoE | MoE |
| Parameters | 1.6T | 550B |
| Context length | 1M | 1M |
| Max output | 384K | Not available |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | DeepSeek-V4-Pro | Nemotron 3 Ultra |
|---|---|---|
| Text input | $0.435 / 1M tokens | Not public |
| Text output | $0.87 / 1M tokens | Not public |
| Cache read | $0.003625 / 1M tokens | Not public |
One or both models have incomplete public pricing.
Summary
- DeepSeek-V4-Proleads in:Coding and Software Engineer (2/3), AI Agent - Information Search (1/1), AI Agent - Tool Usage (1/1)
- Nemotron 3 Ultraleads in:General Knowledge (2/3), Math and Reasoning (1/1)
On average across the 9 shared benchmarks, Nemotron 3 Ultra scores 2.83 higher.
Largest single-benchmark gap: IMO-AnswerBench — DeepSeek-V4-Pro 35.30 vs Nemotron 3 Ultra 92.30 (-57).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.