opensource/cloudpods
3
0
Fork 0
mirror of https://github.com/yunionio/cloudpods.git synced 2026-08-30 17:13:08 +08:00
Code Issues Packages Projects Releases Wiki Activity

feat(aiproxy): move visual provider to model columns and support streaming responses (#25144)

Browse Source 
Promote visual_provider_id/visual_model_key to dedicated ai_model columns,
force non-stream upstream orchestration for visual paths, and synthesize SSE
chunks for streaming Responses requests that include images.
This commit is contained in:
Zexi Li
2026-07-14 18:39:30 +08:00
committed by GitHub
parent 73d7cbbb1f
commit cdc8ab78b5
20 changed files with 817 additions and 97 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/aiproxy/functional-test.md
+1 -1
View File
@@ -301,7 +301,7 @@ startup_timeout_sec = 3600
tool_timeout_sec = 3600
```
**models_catalog.json**（节选）：包含该 virtual key 可见模型的 Codex 元数据（`base_instructions`、`truncation_policy`、`shell_type` 等），避免 Codex 回退到内置 preset。
**models_catalog.json**（节选）：包含该 virtual key 可见模型的 Codex 元数据（`base_instructions`、`truncation_policy`、`shell_type`、`input_modalities` 等），避免 Codex 回退到内置 preset。若绑定 `ai_model` 已启用 Visual（`visual_active`：`config.extensions.visual.enabled` + `visual_provider_id` + `visual_model_key`），对应条目会写入 `input_modalities: ["text","image"]`，使 Codex 按多模态发图。
**aiproxy.env**