why is gemini's default model names like this

[John4650-hub]

i had a bug in my app where if the selected provider is gemini it failed miserably , i spent 1.5 hours just looking at my code thinking about what could be the cause yet the issue was how the default model was named.

"defaultModels": {
        "pollinations": "openai",
        "azure": "model-router3",
        "gemini": "models/gemini-2.5-flash",
        "openrouter": "openai/gpt-oss-120b:free",
        "nvidia": "deepseek-ai/deepseek-v3.2",
        "ollama": "deepseek-v3.2",
        "groq": "openai/gpt-oss-120b",
        "gpt4free.pro": "deepseek-ai/deepseek-v3.2",
        "puter": "gpt-5.2-chat-latest",
        "perplexity": "turbo"
}

when i fetch gemini's models the gemini-2.5-flash model has no models/. This really frustrated me so much that i had to ask for why it's like this?

[hlohaus]

Please consider utilizing the following URL, which is employed by the Gemini live provider:

https://g4f.space/api/gemini-v1beta

There are three distinct custom Gemini providers available: Gemini, GeminiCLI, and v1beta.

Thank you.

[John4650-hub]

Thank you
why isn't this included in the https://g4f.dev/dist/js/providers.json?