DeepSeek: models, pricing and benchmarks
Chinese AI lab known for open-weight reasoning models.
All DeepSeek models
Sorted by average benchmark score. Prices are per million tokens and context is measured in tokens.
| # | Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|---|
| 1 | DeepSeek-V3.2 | DeepSeek | 1000000 | $5.88 | 84.0 |
| 2 | DeepSeek-V3.2 Instruct | DeepSeek | 128000 | $5.91 | 83.7 |
| 3 | DeepSeek-R1 Instruct | DeepSeek | 1000000 | $5.90 | 83.1 |
| 4 | DeepSeek-R1 | DeepSeek | 128000 | $5.83 | 82.8 |
| 5 | DeepSeek-V3.2 Chat | DeepSeek | 1000000 | $5.80 | 82.5 |
| 6 | DeepSeek-R1 Chat | DeepSeek | 400000 | $5.62 | 82.4 |
| 7 | DeepSeek-V3.2 Base | DeepSeek | 128000 | $5.05 | 80.2 |
| 8 | DeepSeek-R1 Base | DeepSeek | 400000 | $5.05 | 80.1 |
| 9 | DeepSeek-Coder | DeepSeek | 32768 | $4.55 | 78.1 |
| 10 | DeepSeek-Coder Instruct | DeepSeek | 32768 | $4.34 | 77.7 |
| 11 | DeepSeek-Coder Chat | DeepSeek | 200000 | $4.20 | 77.3 |
| 12 | DeepSeek-Coder Base | DeepSeek | 32768 | $3.53 | 74.1 |
| 13 | DeepSeek-V3.2 8B | DeepSeek | 32768 | $2.02 | 66.2 |
| 14 | DeepSeek-R1 8B | DeepSeek | 16384 | $1.87 | 64.3 |
| 15 | DeepSeek-V3.2 7B | DeepSeek | 4096 | $1.75 | 63.2 |
| 16 | DeepSeek-R1 7B | DeepSeek | 32768 | $1.79 | 63.0 |
| 17 | DeepSeek-Coder 8B | DeepSeek | 32768 | $1.26 | 58.7 |
| 18 | DeepSeek-Coder 7B | DeepSeek | 32768 | $1.14 | 57.3 |
| 19 | DeepSeek-V3.2 1.5B | DeepSeek | 4096 | $1.02 | 56.9 |
| 20 | DeepSeek-R1 1.5B | DeepSeek | 16384 | $1.04 | 56.6 |
| 21 | DeepSeek-Coder 1.5B | DeepSeek | 16384 | $0.58 | 51.0 |
Best DeepSeek models
DeepSeek-V3.2
DeepSeek-V3.2 is a frontier model from DeepSeek. It offers a 1,000,000-token context window, costs $5.88 per million input tokens and $24.70 per million output tokens, and averages 84.0 across the index's benchmark suite.
frontierDeepSeek-V3.2 Instruct
DeepSeek-V3.2 Instruct is a frontier model from DeepSeek. It offers a 128,000-token context window, costs $5.91 per million input tokens and $29.01 per million output tokens, and averages 83.7 across the index's benchmark suite.
frontierDeepSeek-R1 Instruct
DeepSeek-R1 Instruct is a frontier model from DeepSeek. It offers a 1,000,000-token context window, costs $5.90 per million input tokens and $26.49 per million output tokens, and averages 83.1 across the index's benchmark suite.
multimodalDeepSeek-R1
DeepSeek-R1 is a multimodal model from DeepSeek. It offers a 128,000-token context window, costs $5.83 per million input tokens and $25.23 per million output tokens, and averages 82.8 across the index's benchmark suite.
multimodalDeepSeek-V3.2 Chat
DeepSeek-V3.2 Chat is a multimodal model from DeepSeek. It offers a 1,000,000-token context window, costs $5.80 per million input tokens and $26.43 per million output tokens, and averages 82.5 across the index's benchmark suite.
How DeepSeek fits
DeepSeek has 21 models in the index, of which 21 are open weights. The average input price across the lineup is $3.53 per million tokens, and the average output price is $16.26 per million tokens. The strongest model by average benchmark score is deepseek-v3-2.
Use the table above to compare individual models, or open a model page for its full benchmark breakdown. If you are choosing between providers, start with the comparison pages and then check context limits and licence terms.
FAQ
How many DeepSeek models are in the index?
DeepSeek has 21 models in this index, of which 21 are open weights.
What is the cheapest DeepSeek model?
DeepSeek-Coder 1.5B is the cheapest DeepSeek model by input price at $0.58 per million tokens.
What is the best DeepSeek model?
DeepSeek-V3.2 has the highest average benchmark score in the DeepSeek lineup at 84.0.
Are DeepSeek models open weights?
21 of the 21 DeepSeek models in this index are open weights. The rest are proprietary.
Where is DeepSeek based?
DeepSeek was founded in 2023 and is headquartered in Hangzhou, China.