Flagship Claude model for deep reasoning, coding, and long-horizon agents
Details
| Field | Value |
|---|---|
| id | anthropic/claude-opus-4 |
| family | claude-opus |
| modality | image+text+pdf->text |
| context_length | 200000 |
| max_output_tokens | 32000 |
| release_date | 2025-05-22 |
| is_open_weights | no |
| supports_tool_call | yes |
| supports_reasoning | yes |
| supports_structured_output | no |
| supports_attachment | yes |
| input_modalities | image, text, pdf |
| output_modalities | text |