Model Comparison
Nemotron 3 Super 120B A12B (Reasoning)
vs. Sonar Reasoning Pro
Comparing 2 AI models · 9 benchmarks · NVIDIA, Perplexity
Recommended Pick
Pe Sonar Reasoning Pro 5 metric wins Strongest on: Input price, Output price, TTFT
Lowest Price
Sonar Reasoning Pro
$0.00/1M input price
Best Reasoning
Sonar Reasoning Pro
66.4 reasoning score
Blends available reasoning benchmarks
Best for Coding
Nemotron 3 Super 120B A12B (Reasoning)
31.2 coding index
Composite Indices
Higher is better; speed and price are normalized
Standard Benchmarks
Only benchmarks with data are shown
Differences That Matter
Price gap
Sonar Reasoning Pro is ∞x cheaper on input tokens than Nemotron 3 Super 120B A12B (Reasoning).
Reasoning gap
Sonar Reasoning Pro leads Nemotron 3 Super 120B A12B (Reasoning) by 21.4 points on reasoning.
Top-pick rationale
Sonar Reasoning Pro wins 5 measurable categories, including Input price, Output price, TTFT, Latency.
Response Face-Off
Run one prompt through the selected models and compare response quality with live speed and cost context.
Nemotron 3 Super 120B A12B (Reasoning)
NVIDIA
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Sonar Reasoning Pro
Perplexity
TTFT
—
Time
—
tok/s
—
Tokens
—
Cost
—
Which answer was more useful?
Full Comparison
| Metric | NV Nemotron 3 Super 120B A12B (Reasoning) | Top Pick Pe Sonar Reasoning Pro |
|---|---|---|
| Pricing per 1M tokens | ||
| Input Cost | $0.30/1M | $0.00/1M |
| Output Cost | $0.75/1M | $0.00/1M |
| Blended (3:1) | $0.41/1M | — |
| Specifications | ||
| Organization | NVIDIA | Perplexity |
| Release Date | Mar 11, 2026 | Jan 28, 2025 |
| Performance & Speed | ||
| Throughput | 154.8 tok/s | — |
| TTFT | 964ms | — |
| Latency | 13883ms | — |
| Composite Indices | ||
| Value Score | 100.0 | — |
| Reasoning Score | 45.1 | 66.4 |
| Intelligence | 36.0 | 24.6 |
| Coding | 31.2 | — |
| Standard Benchmarks | ||
| GPQA | 80.0% | — |
| HLE | 19.2% | — |
| MATH 500 | — | 95.7% |
| AIME (Original) | — | 79.0% |
| SciCode | 36.0% | — |
| LCR | 60.0% | — |
| IFBench | 71.5% | — |
| TAU-bench v2 | 67.8% | — |
| TerminalBench Hard | 28.8% | — |
Key Takeaways
Sonar Reasoning Pro offers the best value at $0.00/1M, making it ideal for high-volume applications and cost-conscious projects.
Sonar Reasoning Pro has the strongest reasoning profile with a 66.4 reasoning score, combining the available reasoning-heavy benchmarks.
Nemotron 3 Super 120B A12B (Reasoning) reaches a 31.2 coding index, making it the top choice for software development and code generation tasks.
All models support context windows of ∞+ tokens, suitable for processing lengthy documents and maintaining extended conversations.
When to Choose Each Model
Nemotron 3 Super 120B A12B (Reasoning)
- Code generation
- Software development
Sonar Reasoning Pro
- Cost-sensitive applications
- High-volume processing
- Complex reasoning tasks
- Research & analysis
