更新日: 2026-06-06
gpt-5.5 Vision - URL Method
gpt-5.5 Vision - Local Image
Multi-Image Comparison
Claude Vision
Claude models also support vision through the OpenAI-compatible format:Supported Vision Models
Vision input with
gpt-5.5 and claude-opus-4-8 was verified in Crazyrouter production on March 23, 2026. Image size should not exceed 20MB; base64 encoding adds approximately 33% to the payload size, so URL method is recommended for larger files.