11 models · avg score 72
Track reliability alongside pricing and model depth so provider selection stays grounded in operating reality.
All Systems Operational
Rolling availability across the last 30 days.
Composite reliability indicator from recent status behavior.
Includes degraded performance, partial outages, and major incidents.
| # | Model | Score |
|---|---|---|
| 89 | R1 0528 | 78 |
| 94 | DeepSeek V3.2 Exp | 77 |
| 96 | DeepSeek V3.2 Speciale | 77 |
| 99 | DeepSeek V3.1 Terminus | 77 |
| 115 | DeepSeek V3.2 | 74 |
| 119 | DeepSeek V3.1 | 74 |
| 127 | DeepSeek V3 0324 | 73 |
| 157 | DeepSeek V3 | 70 |
| 168 | R1 | 68 |
| 218 | R1 Distill Llama 70B | 61 |
| 222 | R1 Distill Qwen 32B | 60 |
| Rank | Model | Score |
|---|---|---|
| #89 | R1 0528 | 78 |
| #94 | DeepSeek V3.2 Exp | 77 |
| #96 | DeepSeek V3.2 Speciale | 77 |
| #99 | DeepSeek V3.1 Terminus | 77 |
| #115 | DeepSeek V3.2 | 74 |
| #119 | DeepSeek V3.1 | 74 |
| #127 | DeepSeek V3 0324 | 73 |
| #157 | DeepSeek V3 | 70 |
| #168 | R1 | 68 |
| #218 | R1 Distill Llama 70B | 61 |
| #222 | R1 Distill Qwen 32B | 60 |
DeepSeek currently offers 11 AI models on our platform, with an average quality score of 72/100.
The highest-rated DeepSeek model is R1 0528, ranked #89 overall with a composite score of 78/100.
DeepSeek models have an average score of 72/100. Their best model, R1 0528, is ranked #89 globally. Visit our provider comparison page for detailed analysis.