feat(llm): support Gemini provider

[lloydchang]

fix #161

---

Important

Add support for Gemini provider, including API key handling, model initialization, and documentation updates.

• Provider Support:
  • Add Gemini as a provider in models.py, init.py, and llm/__init__.py.
  • Update init_llm() in llm/__init__.py to initialize Gemini using OpenAI client.
  • Add Gemini API key handling in init.py and llm_openai.py.
• Documentation:
  • Update providers.rst to include Gemini setup instructions.
• Model Handling:
  • Add Gemini model gemini-1.5-flash-latest to models.py with context, max output, and pricing.
  • Update get_recommended_model() and get_summary_model() in models.py to include Gemini.
• Evaluation:
  • Add Gemini model to default models in eval/main.py.

^{This description was created by for 321a31d. It will automatically update as commits are pushed.}

[ellipsis-dev]

👍 Looks good to me! Reviewed everything up to 321a31d in 19 seconds

More details

• Looked at 164 lines of code in 6 files
• Skipped 0 files when reviewing.
• Skipped posting 3 drafted comments based on config settings.

1. gptme/init.py:95

• Draft comment:
  Update the API key prompt to include Gemini, as it is now supported.
• Reason this comment was not posted:
  Confidence changes required: 50%
  The addition of Gemini as a provider is consistent across the codebase. However, the API key prompt and documentation need to be updated to reflect this change.

2. gptme/init.py:105

• Draft comment:
  Update the message to include Gemini, as it is now supported.
• Reason this comment was not posted:
  Confidence changes required: 50%
  The addition of Gemini as a provider is consistent across the codebase. However, the API key prompt and documentation need to be updated to reflect this change.

3. gptme/llm/__init__.py:254

• Draft comment:
  Ensure that the provider detection logic correctly identifies Gemini based on the base URL.
• Reason this comment was not posted:
  Confidence changes required: 50%
  The addition of Gemini as a provider is consistent across the codebase. However, the API key prompt and documentation need to be updated to reflect this change.

Workflow ID: wflow_dqRjxwm5GUkqahW5

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[lloydchang]

Relates to #161 (comment)

[ErikBjare]

Very nice! Thanks :)