[Feature Request]: 通过 Vertex API 添加 gemini-1.5-pro 的接口

[ShinChven]

[Feature Request]: 通过 Vertex API 添加 gemini-1.5-pro 的接口

[Dean-YZG]

我们已经有计划，改造应用中的 provider实体和model实体的关系，来允许用户自定义provider，以及provider提供的models，这样最大限度适配市面上各个服务提供商的模型接口

[ShinChven]

https://github.com/ShinChven/vertex-ai-proxy

Vertex AI 只能用 Json api key，我写了个中继接口，可以临时解决一下 vertex AI token 刷新问题。

可以配置到 Next Chat 里面用。

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

Title: [Feature Request]: Add gemini-1.5-pro interface through Ver

Problem Description

Currently, AI Studio does not give gemini-1.5-pro API permissions, but the vertex API is available.
The API calling methods of vertex API and AI Studio are different. You need to log in with Google Cloud SDK and obtain an access token to access.

I obtained the configuration information through the following script and have used gemini-1.5-pro on Next Chat.

API_ENDPOINT="us-central1-aiplatform.googleapis.com"
PROJECT_ID=<PROJECT_ID>
LOCATION_ID="us-central1"
MODEL_ID="gemini-1.5-pro-preview-0409"
ACCESS_TOKEN="$(gcloud auth print-access-token)"
URL="https://${API_ENDPOINT}/v1/projects/${PROJECT_ID}/locations/${LOCATION_ID}/publishers/google/models/${MODEL_ID}:streamGenerateContent"

echo "URL:"
echo ${URL}
echo "ACCESS TOKEN:"
echo ${ACCESS_TOKEN}
echo "MODEL_ID:"
echo ${MODEL_ID}

The specific configuration method of Next Chat is:

1. Select Google as Model Provider
2. Endpoint Address Fill in URL
3. Fill in ACCESS_TOKEN for API Key
4. API Version delete default
5. Custom Model fill gemini-1.5-pro-preview-0409

Solution Description

I hope Next Chat can better support Vertex API.

Alternatives Considered

No response

Additional Context

No response

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

We already have plans to transform the relationship between the provider entity and the model entity in the application to allow users to customize the provider and the models provided by the provider, so as to maximize the adaptability to the model interfaces of various service providers on the market.