Models
Available Models
Section titled “Available Models”| Model ID | Provider | Input Price | Output Price | Context | Best For |
|---|---|---|---|---|---|
deepseek-chat | DeepSeek | $0.14 / 1M | $0.28 / 1M | 128K | General purpose, coding, analysis |
deepseek-reasoner | DeepSeek | $0.55 / 1M | $2.19 / 1M | 128K | Reasoning, math, logic (R1) |
qwen-max | Alibaba | $0.50 / 1M | $1.50 / 1M | 32K | Multilingual, translation |
qwen-plus | Alibaba | $0.20 / 1M | $0.60 / 1M | 128K | Balanced price-performance |
glm-4 | Zhipu | $0.10 / 1M | $0.10 / 1M | 128K | Chinese-optimized |
Model Selection Guide
Section titled “Model Selection Guide”DeepSeek V3 (deepseek-chat)
Section titled “DeepSeek V3 (deepseek-chat)”Our most popular model. Excellent for:
- General conversation and chat
- Code generation and debugging
- Content writing and summarization
- Data analysis
DeepSeek R1 (deepseek-reasoner)
Section titled “DeepSeek R1 (deepseek-reasoner)”Reasoning specialist. Best for:
- Complex math problems
- Logic puzzles
- Multi-step reasoning
- Problem-solving
Qwen-Max (qwen-max)
Section titled “Qwen-Max (qwen-max)”Alibaba’s flagship model. Best for:
- Multilingual tasks
- English-Chinese translation
- Document understanding
- Long-form content
GLM-4 (glm-4)
Section titled “GLM-4 (glm-4)”Zhipu’s model optimized for:
- Chinese language scenarios
- Enterprise applications
- Cost-sensitive workloads
Model Aliases
Section titled “Model Aliases”For OpenAI SDK compatibility, you can use these aliases:
| Alias | Maps To |
|---|---|
gpt-4 | deepseek-chat |
gpt-3.5-turbo | qwen-plus |