DeepSeek
36 models from DeepSeek, ranked by Podium Score.
Chinese AI lab known for high-performance open-weight models at competitive prices.
Deepseek V4 Pro High Preview
Top model
73.0
Podium Score
47.2
Average score
21
open weights
DeepSeek API Pricing — per 1M Tokens
Get API Key →| Model | In $/M | Out $/M | Podium Score | Benchmarks |
|---|---|---|---|---|
| $0.27/M | $1.1/M | 73.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 72.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 71.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 71.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 71.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 71.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 71.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 71.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 71.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 71.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 66.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 65.0 | Benchmarks → | |
| $0.44/M | $1.32/M | 65.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 63.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 59.0 | Benchmarks → | |
| $0.3/M | $1.2/M | 54.5 | Benchmarks → | |
| $0.3/M | $1.2/M | 50.0 | Benchmarks → | |
| $0.43/M | $0.87/M | 46.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 46.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 45.3 | Benchmarks → | |
| $0.27/M | $1.1/M | 43.7 | Benchmarks → | |
| $0.27/M | $1.1/M | 42.1 | Benchmarks → | |
| $0.14/M | $0.28/M | 42.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 39.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 38.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 31.9 | Benchmarks → | |
| $0.27/M | $1.1/M | 31.1 | Benchmarks → | |
| $0.27/M | $1.1/M | 29.3 | Benchmarks → | |
| $0.27/M | $1.1/M | 25.9 | Benchmarks → | |
| $0.27/M | $1.1/M | 21.7 | Benchmarks → | |
| $0.27/M | $1.1/M | 21.0 | Benchmarks → | |
| $0.27/M | $1.1/M | 18.6 | Benchmarks → | |
| $0.27/M | $1.1/M | 15.2 | Benchmarks → | |
| $0.27/M | $1.1/M | 12.9 | Benchmarks → | |
| $0.27/M | $1.1/M | 9.8 | Benchmarks → | |
| $0.27/M | $1.1/M | 5.4 | Benchmarks → |
Pricing and speed from official API documentation; scores aggregated from five leaderboards.
Deepseek V4 Pro High Preview
73.073.0
Deepseek V4 Flash High Preview
72.072.0
Deepseek V3.2 Exp Thinking
71.071.0
Deepseek V3.2 Thinking
71.071.0
Deepseek V3.2 Exp
71.071.0
Deepseek R1 0528
71.071.0
Deepseek V3.1
71.071.0
Deepseek V3.1 Terminus Thinking
71.071.0
Deepseek V3.1 Thinking
71.071.0
Deepseek V3.1 Terminus
71.071.0
Deepseek V2.5 1210
66.066.0
Deepseek V2.5
65.065.0
DeepSeek V4 Flash Vision
65.065.0
Deepseek Coder V2
63.063.0
Deepseek LLM 67b Chat
59.059.0
DeepSeek V4.1 Flash
54.554.5
DeepSeek V4.1 Flash (Non-reasoning)
50.050.0
DeepSeek V4 Pro
46.046.0
DeepSeek V4 Flash 0420 (Non-reasoning)
46.046.0
Deepseek V4 Pro 0424
45.345.3
Deepseek V4 Pro 0424 High
43.743.7
Deepseek V4 Flash 0420
42.142.1
DeepSeek V4 Flash
42.042.0
Deepseek V4 Flash 0420 High
39.039.0
Deepseek V3 2 Reasoning
38.038.0
Deepseek V4 Pro 0424 Non Reasoning
31.931.9
Deepseek V3 1 Terminus Reasoning
31.131.1
Deepseek V4 Flash Non Reasoning
29.329.3
Deepseek V3 2 Reasoning 0925
25.925.9
Deepseek V3 2 0925
21.721.7
Deepseek V3 1 Reasoning
21.021.0
Deepseek R1 0120
18.618.6
Deepseek V3 0324
15.215.2
DeepSeek V3.2
12.912.9
Deepseek R1 Distill Llama 70b
9.89.8
DeepSeek R1
5.45.4