{"slug": "gemini-4-hydrogen-reaches-firebase-ai-logic-and-rejects-old-settings", "title": "Gemini 4 Hydrogen Reaches Firebase AI Logic and Rejects Old Settings", "summary": "Firebase AI Logic apps can now call Google's gemini-4-hydrogen model, but Firebase's Nov. 12, 2026 release notes warn that requests setting thinkingBudget fail with a 400 error and that the default thinking level rises from MEDIUM on gemini-3.8-flash to HIGH on Hydrogen, which Firebase says may raise latency and token use unless apps set thinkingLevel to MEDIUM or LOW. Explicit context caching is no longer supported, temperature, topP and topK are ignored, frequencyPenalty, presencePenalty and candidateCount fail with a 400 error, and Google's Nov. 12 Gemini API changelog prices Hydrogen at $1.50 per 1 million input tokens and $7.50 per 1 million output tokens on the paid tier, with thinking tokens billed as output. Hydrogen is labeled a short-term availability model on the Agent Platform Gemini API (formerly Vertex AI), and Firebase advises serving the model name from Remote Config or server prompt templates rather than hard-coding it in client apps.", "body_md": "# Gemini 4 Hydrogen Reaches Firebase AI Logic and Rejects Old Settings\n\nRemove thinkingBudget and explicit caching before switching a Firebase AI Logic app to Gemini 4 Hydrogen, or its requests fail with a 400 error.\n\nFirebase AI Logic apps can now call Google’s new `gemini-4-hydrogen` model, but an app that only swaps the model name may start failing. Requests that set `thinkingBudget` now return a 400 error, and the default thinking level rises from medium to high. Firebase listed the changes in its Nov. 12 release notes ([Firebase release notes, Nov. 12, 2026](https://firebase.google.com/support/releases)), the same day Google made the model generally available in the Gemini API.\n\n## Key Takeaways\n\n- `thinkingBudget` is not supported on Hydrogen. Firebase said there is “no backwards-compatibility” and the request fails with a 400 error.\n- The default thinking level is `HIGH` , up from`MEDIUM` on`gemini-3.8-flash` . Firebase said that may raise latency and token use unless the app sets`thinkingLevel` to`MEDIUM` or`LOW` .\n- Explicit context caching is gone. Only implicit caching runs.\n- On the Agent Platform Gemini API (formerly Vertex AI), Hydrogen is a short-term availability model.\n\n## What Changes\n\nFirebase said Hydrogen works from client apps through both the Gemini Developer API, including its free tier, and the Agent Platform Gemini API. Its notes for teams moving off `gemini-3.8-flash` or other Gemini 3.x models list these differences:\n\n- Thinking: AI Logic supports `LOW` ,`MEDIUM` and`HIGH` . The new`XHIGH` level that Google offers in the Gemini API is not available through AI Logic yet, and`MINIMAL` still fails with a 400 error, as it did on 3.8 Flash.\n- Caching: implicit caching is on by default for prompts of at least 4,096 input tokens on the Gemini Developer API and 6,144 on the Agent Platform API. Explicit caching is not supported.\n- Limits: the input limit of 1,048,576 tokens and the 65,536-token output limit match 3.8 Flash. AI Logic does not yet support Hydrogen’s “long decode” mode for longer outputs.\n- Parameters: `temperature` ,`topP` and`topK` are ignored.`frequencyPenalty` ,`presencePenalty` and`candidateCount` fail with a 400 error.\n- Files: Hydrogen and other Gemini 4.x models accept Microsoft Office and OpenDocument files in addition to PDFs and plain text.\n\nGoogle’s Nov. 12 Gemini API changelog prices Hydrogen at $1.50 per 1 million input tokens and $7.50 per 1 million output tokens on the paid tier. Google’s migration guide says thinking tokens are billed as output. It also tells teams to “test `medium` and `high` on your evaluation suite to balance speed and cost.”\n\nNetlify added the model to its AI Gateway and Agent Runners the same day, its Nov. 12 changelog said, so Netlify Functions can call it without managing API keys.\n\n## The Take\n\nThe 400 errors show up in the first test run. The thinking default is harder to spot: an app that never set a thinking level was getting medium on 3.8 Flash and gets high on Hydrogen, with no code change and no error. Because Google bills thinking tokens as output, the change shows up as slower replies and a larger bill, not as a failed request. Set `thinkingLevel` explicitly before switching, then compare the two levels on production prompts.\n\nThe short-term availability label matters most for mobile teams. A model name hard-coded in an app on the Agent Platform API can be retired faster than users update the app. Firebase’s own advice is to serve the name from Remote Config or server prompt templates. This is the third Gemini model change Firerun has covered in two weeks, after the Nov. 17 speech preview shutdown and the Dec. 1 Gemini 2.5 reroute. Teams still shipping the model name in client code should move it out now.", "url": "https://wpnews.pro/news/gemini-4-hydrogen-reaches-firebase-ai-logic-and-rejects-old-settings", "canonical_source": "https://firerun.io/firebase-ai-logic-gemini-4-hydrogen-upgrade-2026/", "published_at": "2026-11-15 00:00:00+00:00", "updated_at": "2026-11-15 10:51:22.870037+00:00", "lang": "en", "topics": ["large-language-models", "generative-ai", "ai-products", "ai-tools", "developer-tools"], "entities": ["Google", "Firebase AI Logic", "gemini-4-hydrogen", "gemini-3.8-flash", "Gemini API", "Agent Platform Gemini API", "Netlify", "Firerun"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/gemini-4-hydrogen-reaches-firebase-ai-logic-and-rejects-old-settings", "markdown": "https://wpnews.pro/news/gemini-4-hydrogen-reaches-firebase-ai-logic-and-rejects-old-settings.md", "text": "https://wpnews.pro/news/gemini-4-hydrogen-reaches-firebase-ai-logic-and-rejects-old-settings.txt", "jsonld": "https://wpnews.pro/news/gemini-4-hydrogen-reaches-firebase-ai-logic-and-rejects-old-settings.jsonld"}}