--- title: "使用 Vision 模型" url: "/docs/completion/vision" canonical_url: "https://docs.litellm.ai/docs/completion/vision" type: "docs" last_updated: "2026-10-08" related: - "/docs/guides/multimodal_io" - "/docs/completion/audio" --- # 使用 Vision 模型 > Index of all LiteLLM docs: https://docs.litellm.ai/llms.txt ## 快速入門 {#quick-start} 將圖片傳遞給模型的範例 **LiteLLMPython SDK** ```python import os from litellm import completion os.environ["OPENAI_API_KEY"] = "your-api-key" # openai call response = completion( model = "gpt-5.6-terra", messages=[ { "role": "user", "content": [ { "type": "text", "text": "What’s in this image?" }, { "type": "image_url", "image_url": { "url": "https://awsmp-logos.s3.amazonaws.com/seller-xw5kijmvmzasy/c233c9ade2ccb5491072ae232c814942.png" } } ] } ], ) ``` **LiteLLM Proxy Server** 1. 在 config.yaml 上定義 vision 模型 ```yaml model_list: - model_name: gpt-5.6-terra # OpenAI gpt-5.6-terra litellm_params: model: openai/gpt-5.6-terra api_key: os.environ/OPENAI_API_KEY - model_name: llava-hf # Custom OpenAI compatible model litellm_params: model: openai/llava-hf/llava-v1.6-vicuna-7b-hf api_base: http://localhost:8000 api_key: fake-key model_info: supports_vision: True # set supports_vision to True so /model/info returns this attribute as True ``` 2. 執行 proxy server ```bash litellm --config config.yaml ``` 3. 使用 OpenAI Python SDK 測試它 ```python import os from openai import OpenAI client = OpenAI( api_key="sk-", # your litellm proxy api key ) response = client.chat.completions.create( model = "gpt-5.6-terra", # use model="llava-hf" to test your custom OpenAI endpoint messages=[ { "role": "user", "content": [ { "type": "text", "text": "What’s in this image?" }, { "type": "image_url", "image_url": { "url": "https://awsmp-logos.s3.amazonaws.com/seller-xw5kijmvmzasy/c233c9ade2ccb5491072ae232c814942.png" } } ] } ], ) ``` ## 檢查模型是否支援 `vision` {#checking-if-a-model-supports-vision} **LiteLLM Python SDK** 使用 `litellm.supports_vision(model="")` -> 若模型支援 `True` 則回傳 `vision`，若不支援則回傳 `False` ```python keep-model-ids assert litellm.supports_vision(model="openai/gpt-5.6-terra") == True assert litellm.supports_vision(model="vertex_ai/gemini-3.1-pro-preview") == True assert litellm.supports_vision(model="openai/gpt-3.5-turbo") == False assert litellm.supports_vision(model="xai/grok-2-vision-latest") == True assert litellm.supports_vision(model="xai/grok-2-latest") == False ``` **LiteLLM Proxy Server** 1. 在 config.yaml 上定義 vision 模型 ```yaml model_list: - model_name: gpt-5.6-terra # OpenAI gpt-5.6-terra litellm_params: model: openai/gpt-5.6-terra api_key: os.environ/OPENAI_API_KEY - model_name: llava-hf # Custom OpenAI compatible model litellm_params: model: openai/llava-hf/llava-v1.6-vicuna-7b-hf api_base: http://localhost:8000 api_key: fake-key model_info: supports_vision: True # set supports_vision to True so /model/info returns this attribute as True ``` 2. 執行 proxy server ```bash litellm --config config.yaml ``` 3. 呼叫 `/model_group/info` 以檢查您的模型是否支援 `vision` ```shell curl -X 'GET' \ 'http://localhost:4000/model_group/info' \ -H 'accept: application/json' \ -H "x-api-key: $LITELLM_API_KEY" ``` 預期回應 ```json { "data": [ { "model_group": "gpt-5.6-terra", "providers": ["openai"], "max_input_tokens": 922000, "max_output_tokens": 128000, "mode": "chat", "supports_vision": true, # 👈 supports_vision is true "supports_function_calling": true }, { "model_group": "llava-hf", "providers": ["openai"], "max_input_tokens": null, "max_output_tokens": null, "mode": null, "supports_vision": true, # 👈 supports_vision is true "supports_function_calling": false } ] } ``` ## 明確指定影像類型 {#explicitly-specify-image-type} 如果您有沒有 mime-type 的圖片，或是 litellm 錯誤地推斷了您圖片的 mime type（例如在 vertex ai 中呼叫 `gs://` 的網址），您可以透過 `format` 參數明確設定。 ```json "image_url": { "url": "gs://my-gs-image", "format": "image/jpeg" } ``` LiteLLM 會將其用於任何支援指定 mime-type 的 API 端點（例如 anthropic/bedrock/vertex ai）。 對於其他端點（例如 openai），則會忽略。 **SDK** ```python import os from litellm import completion os.environ["ANTHROPIC_API_KEY"] = "your-api-key" # openai call response = completion( model = "claude-sonnet-5", messages=[ { "role": "user", "content": [ { "type": "text", "text": "What’s in this image?" }, { "type": "image_url", "image_url": { "url": "https://awsmp-logos.s3.amazonaws.com/seller-xw5kijmvmzasy/c233c9ade2ccb5491072ae232c814942.png", "format": "image/jpeg" } } ] } ], ) ``` **PROXY** 1. 在 config.yaml 上定義 vision 模型 ```yaml model_list: - model_name: gpt-5.6-terra # OpenAI gpt-5.6-terra litellm_params: model: openai/gpt-5.6-terra api_key: os.environ/OPENAI_API_KEY - model_name: llava-hf # Custom OpenAI compatible model litellm_params: model: openai/llava-hf/llava-v1.6-vicuna-7b-hf api_base: http://localhost:8000 api_key: fake-key model_info: supports_vision: True # set supports_vision to True so /model/info returns this attribute as True ``` 2. 執行 proxy server ```bash litellm --config config.yaml ``` 3. 使用 OpenAI Python SDK 測試它 ```python import os from openai import OpenAI client = OpenAI( api_key="sk-", # your litellm proxy api key ) response = client.chat.completions.create( model = "gpt-5.6-terra", # use model="llava-hf" to test your custom OpenAI endpoint messages=[ { "role": "user", "content": [ { "type": "text", "text": "What’s in this image?" }, { "type": "image_url", "image_url": { "url": "https://awsmp-logos.s3.amazonaws.com/seller-xw5kijmvmzasy/c233c9ade2ccb5491072ae232c814942.png", "format": "image/jpeg" } } ] } ], ) ``` ## 規格 {#spec} ``` "image_url": str OR "image_url": { "url": "url OR base64 encoded str", "detail": "openai-only param", "format": "specify mime-type of image" } ``` ## Related pages - [Multimodal I/O](https://docs.litellm.ai/docs/guides/multimodal_io.md) - [使用 Audio 模型](https://docs.litellm.ai/docs/completion/audio.md)