Author's Description
Solar Pro 3 is Upstage's powerful Mixture-of-Experts (MoE) language model. With 102B total parameters and 12B active parameters per forward pass, it delivers exceptional performance while maintaining computational efficiency. Optimized for Korean with English and Japanese support.
Key Specifications
Supported Parameters
This model supports the following parameters:
Features
This model supports the following features:
Performance Summary
Upstage's Solar Pro 3, a 102B parameter Mixture-of-Experts model with 12B active parameters, demonstrates competitive response times, ranking in the 54th percentile for speed. It also offers cost-effective solutions, placing in the 67th percentile for price. A standout feature is its exceptional reliability, achieving a perfect 100% success rate across all benchmarks, indicating consistent and usable responses. The model exhibits strong performance in General Knowledge (95.5% accuracy) and Ethics (99.0% accuracy), showcasing its ability to handle broad factual queries and ethical dilemmas effectively. Its Mathematics (85.0% accuracy) and Coding (82.0% accuracy) capabilities are also solid. However, Solar Pro 3 shows notable weaknesses in Hallucinations, with only 52.0% accuracy in acknowledging uncertainty, and Instruction Following (50.0% accuracy), suggesting areas for improvement in handling ambiguous or complex directives. Its Email Classification accuracy (92.0%) is decent but ranks lower compared to other models.
Model Pricing
Current Pricing
| Feature | Price (per 1M tokens) |
|---|---|
| Prompt | $0.15 |
| Completion | $0.6 |
| Input Cache Read | $0.015 |
Price History
Available Endpoints
| Provider | Endpoint Name | Context Length | Pricing (Input) | Pricing (Output) |
|---|---|---|---|---|
|
Upstage
|
Upstage | upstage/solar-pro-3 | 128K | $0.15 / 1M tokens | $0.6 / 1M tokens |
Benchmark Results
| Benchmark | Category | Reasoning | Strategy | Free | Executions | Accuracy | Cost | Duration |
|---|