fix(api-nodes): add support for "thought_image" in Nano Banana 2

[bigcat88]

And corrected price badges

API Node PR Checklist

Scope

• Is API Node Change

Pricing & Billing

• Need pricing update
• No pricing update

If Need pricing update:

• Metronome rate cards updated
• Auto‑billing tests updated and passing

QA

• QA done
• QA not required

Comms

• Informed Kosinkadink

[coderabbitai]

📝 Walkthrough

Walkthrough

The pull request extends image generation nodes to support thinking context visualization. The get_image_from_response method has been updated to accept a thought parameter for filtering image parts based on thinking association. Both the GeminiImage and GeminiNanoBanana2 nodes now expose an additional output for thought-derived images. Thinking configuration is integrated into the image generation requests for these nodes. As a result, both nodes' execute methods return three outputs instead of two, enabling access to images generated from the model's thinking process.

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 25.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly describes the primary change: adding support for 'thought_image' in Nano Banana 2, which matches the main objective of exposing thinking-aware image outputs.
Description check	✅ Passed	The pull request description directly addresses the changeset by mentioning 'thought_image' support in Nano Banana 2 and price badge corrections, which align with the file-level changes.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

📝 Coding Plan

• Generate coding plan for human review comments

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@comfy_api_nodes/nodes_gemini.py`:
- Around line 936-940: The tooltip for IO.Image.Output ("First image from the
model's thinking process") is inaccurate because
get_image_from_response(response, thought=True) may return multiple images;
update the tooltip text in the IO.Image.Output instantiation to reflect plural
results (e.g., "Thought image(s) from the model's thinking process; may return
multiple images") so it matches the behavior of get_image_from_response;
reference IO.Image.Output, get_image_from_response, and the node's execute
implementation when making the change.
- Around line 65-69: The Flash price literals in the pricing map ($flashPrices)
are higher than official Gemini 3.1 Flash Image rates; update the $flashPrices
values to match Google Cloud docs (1k: 0.067, 2k: 0.101, 4k: ~0.151) and keep
the rest of the logic unchanged (the $isFlash flag, $proPrices map, $prices
selection, and the final return formatting should remain as-is).

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 6e3e7e4a-e7d3-4f73-bd29-4b45fa23ab09

📥 Commits

Reviewing files that changed from the base of the PR and between b941913 and 29fd344.

⛔ Files ignored due to path filters (1)

• comfy_api_nodes/apis/gemini.py is excluded by !comfy_api_nodes/apis/**

📒 Files selected for processing (1)

• comfy_api_nodes/nodes_gemini.py