No Image Generation Models Available After Gemini CLI Setup

[danielnguyenkarate]

Bug Report: No Image Generation Models Available After Gemini CLI Setup

Is it a request payload issue?

• Yes, this is a request payload issue.
• No, it's another issue.

---

Describe the bug

After setting up Gemini CLI according to Tutorial-2, I cannot find or use any image generation models.

The documentation (Tutorial-3) states that:

"Gemini CLI now fully covers Gemini Web and is completely free"

However, in practice, after completing the setup in Tutorial-2, only text models are available. No image-capable models (such as NanoBanana or Gemini image models) can be found or used.

There is no clear way to generate images using Gemini CLI alone.

---

CLI Type

• gemini-cli

---

Model Name

• gemini-2.5-pro (text only)
• ❌ No image generation models available

---

LLM Client

• CLIProxyAPI (OpenAI-compatible API)
• Tested via cURL / Cherry Studio

---

Request Information

This is not a request payload issue.

Even before sending any request, there are no image generation models available to select or call.

Gemini CLI authentication was completed successfully:

• OAuth login succeeded
• Text generation works as expected
• Image generation models are missing

---

Expected behavior

After setting up Gemini CLI and based on the notice that Gemini CLI fully covers Gemini Web, I expected:

• Image generation models to be available via Gemini CLI
  OR
• Clear documentation stating that image generation is not yet supported in Gemini CLI and still requires Gemini Web (CLIProxyAPI v6.1.X)

---

Actual behavior

• Gemini CLI works only for text models
• No image generation models are available
• Tutorial-3 still relies on Gemini Web cookies (v6.1.X), which contradicts the claim that Gemini CLI fully covers Gemini Web

---

Screenshots

N/A (no image models available to display)

---

OS Type

• OS: Windows
• Version: Windows 10 / 11 (x64)

---

Additional context

• Followed Tutorial-2 exactly:
  https://help.router-for.me/hands-on/tutorial-2.html
• Tutorial-3 still depends on Gemini Web cookies:
  https://help.router-for.me/hands-on/tutorial-3
• The statement about Gemini CLI fully covering Gemini Web is confusing, as image generation still appears unavailable without Gemini Web support.

[Galang23]

Gemini CLI alone does not provide image generation model. Login with Antigravity to use it, and send request to chat completions endpoint.

[danielnguyenkarate]

Gemini CLI alone does not provide image generation model. Login with Antigravity to use it, and send request to chat completions endpoint.

I have used Antigravity but the quotas is so worst. I cannot have "Gemini Images Free"

[hungthai1401]

try to use AIStudio

[danielnguyenkarate]

try to use AIStudio

Is there any hand-ons tutorial?
I have followed the instruction in https://help.router-for.me/configuration/provider/ai-studio.html but I don't understand? It still doesn't have image generate models? Seems like the instruction want to make the CLIProxyAPI as a backend rather than create another endpoint for images model