How do I add a model to the gemini models list ?

[CeritAli]

Hello, I want to add 2.5 flash but I cannot add it as a gemini model (it is not present in the gemini models list). I can only add it as an open-ai compatible api but this one adds a lot of latency which makes it unusable for production. I have limited knowledge so I would prefer if there is a simple way or if you could explain to me step by step. Thank you guys for your help!

[dosubot[bot]]

Hello @CeritAli! I'm Dosu, a friendly bot here to help you with bugs, answer your questions, and assist you in becoming a contributor. I'm here while you wait for a human maintainer. How can I assist you today?

To add a new model to the Gemini models list, follow these steps:

1. Update the Configuration File:

   • Open the conf/llm_factories.json file.

   • Add a new entry for your model under the "Gemini" section. Here is an example configuration:

     {
         "llm_name": "gemini-2.5-flash",
         "tags": "LLM,CHAT,1024K",
         "max_tokens": 1048576,
         "model_type": "chat",
         "is_tools": true
     }

   • Adjust the llm_name, tags, max_tokens, model_type, and is_tools fields as necessary to match the specifications of your new model ^[1].

2. Use Existing Scripts:

   • The add_llm function in the api/apps/llm_app.py file can be used to add new models. This function handles various model factories and their specific authentication methods, constructs the necessary API keys, and saves the model configuration using the TenantLLMService.save method ^[2].

3. Front-End Interface:

   • If you prefer using a graphical interface, the front-end code in web/src/pages/user-setting/setting-model/index.tsx provides a user interface for adding models. The handleAddModel function determines the appropriate modal to display based on the selected model factory ^[3].

By following these steps, you should be able to add your model to the Gemini models list and avoid the latency issues associated with using it as an OpenAI-compatible API.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}