model parameter to switch between them:
Anthropic / Claude
| Model ID | Description | Vision | Thinking |
|---|---|---|---|
claude-haiku-4-5 | Fastest and most cost-effective Claude model, ideal for high-throughput lightweight tasks | ✓ | |
claude-sonnet-4-6 | Best balance of performance and cost, suitable for most business use cases | ✓ | |
claude-opus-4-6 | Claude flagship model, ideal for complex reasoning and high-quality generation | ✓ |
OpenAI / GPT-5
| Model ID | Description | Vision | Thinking |
|---|---|---|---|
gpt-5 | OpenAI flagship model with powerful general reasoning and multimodal capabilities | ✓ | |
gpt-5-codex | GPT-5 optimized for code — excels at code generation and debugging | ||
gpt-5-codex-mini | Lightweight code model with ultra-fast response and low cost | ||
gpt-5.1 | GPT-5 enhanced, with improved instruction following and reasoning | ✓ | |
gpt-5.1-codex | GPT-5.1 code-specialized variant | ||
gpt-5.1-codex-max | GPT-5.1 flagship code model, maximum coding capability | ||
gpt-5.1-codex-mini | GPT-5.1 lightweight code variant, fast and cost-efficient | ||
gpt-5.2 | GPT-5 second generation with enhanced multimodal capabilities | ✓ | |
gpt-5.2-codex | GPT-5.2 code-specialized variant | ||
gpt-5.3-codex | Latest GPT-5 code model with continuously improving capabilities | ||
gpt-5.3-codex-spark | GPT-5.3 Codex Spark variant, focused on high-speed code generation | ||
gpt-5.4 | GPT-5 generation 4, enhanced general reasoning and multimodal capabilities | ✓ |
Google / Gemini
| Model ID | Description | Vision | Thinking |
|---|---|---|---|
gemini-3.1-fast | Google Gemini fast variant, high-value multimodal model | ✓ | |
gemini-3.1-thinking | Google Gemini thinking variant with built-in reasoning chain, ideal for complex problem solving | ✓ | ✓ |
DeepSeek
| Model ID | Description | Vision | Thinking |
|---|---|---|---|
deepseek-v3.2 | DeepSeek flagship model with extended thinking support, excels at code and complex reasoning | ✓ |
Alibaba / Qwen
| Model ID | Description | Vision | Thinking |
|---|---|---|---|
qwen3-max | Qwen flagship model with deep thinking support, excels at complex reasoning and code generation | ✓ | |
qwen3.5-plus | Qwen enhanced model, balanced performance and cost | ||
qwen3.5-flash | Qwen ultra-fast model, ideal for low-latency and high-throughput scenarios |
Zhipu / GLM
| Model ID | Description | Vision | Thinking |
|---|---|---|---|
glm-5 | Zhipu flagship model with outstanding Chinese language understanding and generation |
Moonshot / Kimi
| Model ID | Description | Vision | Thinking |
|---|---|---|---|
kimi-k2.5 | Moonshot flagship model with ultra-long context and strong Chinese language capabilities |
MiniMax
| Model ID | Description | Vision | Thinking |
|---|---|---|---|
MiniMax-M2.5 | MiniMax flagship model with excellent cost-effectiveness |