wavespeed-ai/any-llm/vision

Positive prompt words for content generation (required).

System prompt (optional), used to provide context or instructions for the model.

List of pending image URLs (optional).

Whether to include reasoning content in the final answer (optional).

Generate priority (optional). throughput is suitable for most scenarios; latency is suitable for low-latency scenarios.

Sampling temperature (optional). Lower values produce more stable output, while higher values result in more diverse output.

Maximum number of generated tokens (optional). The actual limit is the context length minus the prompt length.

Model name used (optional). Advanced models are billed at 3× rate: google/gemini-pro-1.5, openai/gpt-4.1, openai/gpt-4o, anthropic/claude-3.5-sonnet, openai/gpt-5-chat, meta-llama/llama-3.2-90b-vision-instruct, anthropic/claude-3.7-sonnet.

Enable synchronous mode (optional). When set to true, the function will wait for generation and upload to complete before returning a response. Available only via API.

Enable BASE64 output (optional). When enabled, the result will be returned as a BASE64 string instead of a URL, available only via API.

Callback interface URL (optional). After the task is generated, the system will send a result request to this link. A status code 200 indicates successful receipt.