Reported News
Google launches Gemini 4 Hydrogen, a fast agentic model at $1.50 / $7.50 per 1M tokens for everyone
Gemini 4 Hydrogen is Google's fast Gemini 4 model, released Nov 12, 2026 for "rapid execution across diverse agentic workflows." Google prices it flat at $1.50 per 1M input and $7.50 per 1M output tokens across a 1M-token context, and is rolling it out to the Gemini API, AI Studio, the Gemini app and Search.
- Changes reported
- 6
- listed below, each from a source we read
- We tested
- Not tested yet
- reported from sources; we did not run it ourselves
- Gaps we flag
- 4
- things we could not verify
- Sources
- 2
- 2 primary, linked below
What changed
| # | What the source says | Status |
|---|---|---|
| 01 | Google released Gemini 4 Hydrogen (API model code gemini-4-hydrogen), a fast model for everyday coding and agent work that sits below the frontier Gemini 4 Carbon and Argon, and is meant to run alone or as the worker model beside them. | Reported |
| 02 | Price, per the post: a flat $1.50 per 1M input tokens and $7.50 per 1M output tokens across the whole 1M-token context. The API docs list a 1,048,576-token context, 65,536 default max output, up to 1,048,576 with long decode, and text, image, video, audio and PDF input. | Reported |
| 03 | Four thinking levels: low, medium, high (the default) and a new xhigh tier. Google says Hydrogen scores 51 on the Artificial Analysis Intelligence Index at high (Argon: 50) and 54 at xhigh. | Reported |
| 04 | Against Gemini 3.8 Flash, Google claims 2.5x faster task completion at the default setting, 57% vs 18.2% on Terminal-Bench 4.1, and more than 5x fewer fabricated claims. | Reported |
| 05 | Google Antigravity gets Orchestrator, which switches between Carbon and Hydrogen within one session; Google says it reached near-Carbon success on Terminal-Bench 4.0 with 24% less quota and 19% faster completion on average. | Reported |
| 06 | Rollout starts today for all AI Studio and Gemini API tiers, Gemini for business, the Gemini app model menu and Google Search; Antigravity access is for paid AI Pro and Ultra subscribers, and free app users get it only through auto model selection. | Reported |
reported by the sourcemeasured or run by usnot verified
Primary source: Google: Introducing Gemini 4 Hydrogen. We are the fast, independent index; the vendor’s own page is the source of record for this release.
What the price means next to Flash and Carbon
At $1.50 / $7.50 per 1M tokens, Hydrogen costs the same as the list price Google has scheduled for Gemini 3.8 Flash from Jan 1, 2027, and twice the $0.75 / $3.75 that Flash is billed at through Dec 31, 2026 (per Google's pricing page as of Nov 1). It is below Carbon's introductory $2 / $10 and well below Carbon's $4 / $20 list price.
So for a Flash user the real question for the next seven weeks is whether Hydrogen finishes enough work per token to beat a model that is temporarily half its price. Google's own partner figures (2x to 2.8x faster and 12% to 23% fewer tokens than 3.8 Flash) would not by themselves cover a 2x per-token price gap; reliability gains and fewer retries would have to make up the rest. After Jan 1 the per-token prices match.
| Model | Input | Output | Note |
|---|---|---|---|
| Gemini 4 Hydrogen | $1.50 | $7.50 | Flat across 1M context |
| Gemini 3.8 Flash (to Dec 31) | $0.75 | $3.75 | Rises to $1.50 / $7.50 on Jan 1, 2027 |
| Gemini 4 Carbon (intro) | $2 | $10 | $4 / $20 after Nov 30, 2026 |
Questions Google's post leaves open
- Every benchmark is Google's own run or a third-party leaderboard Google cites; none of the scores comes with the exact settings or the competing models it calls its class.
- The API page says context caching is implicit only and explicit caching is not supported, so long repeated prompts may not get the cached-input discount Carbon advertises.
- The post does not say whether the 3.8 Flash promo price or the model itself will be retired once Hydrogen is the default.
In the source’s words
“Hydrogen is built for rapid execution across diverse agentic workflows.”
“Gemini 4 Hydrogen is priced flat across its entire 1 million-token context window at $1.50 per million input tokens and $7.50 per million output tokens.”
“Orchestrator dynamically switches between models like Carbon and Hydrogen within the same session.”
What we have not verified
- We did not run Gemini 4 Hydrogen; this brief is based on Google's post and API docs.
- All benchmark, speed, token-saving and safety figures are Google's or its partners' claims; we did not reproduce them.
- Pricing for cached input, batch use or the xhigh thinking level beyond the flat rate is not stated in the post.
- Global availability timing per country and per product surface was not checked.
This brief is Reported: written from the sources below, not from running the release ourselves.
Sources
- Google: Introducing Gemini 4 Hydrogenprimary source · fetched 2026-11-12
- Gemini API docs: Gemini 4 Hydrogen model pageprimary source · fetched 2026-11-12
Related
Spotted an error? See our corrections policy or email the address on the contact page. Markdown version for agents: /news/google-gemini-4-hydrogen.md
Want these briefs in your Google Top Stories? Add Agentic Weekly as a preferred source on Google.
Newsletter
The Weekly Verdict. Free, every Monday.
A free email every Monday: the week's few big AI releases, then 5 short use cases for the new models, each with a prompt you can paste.
Free. Double opt-in. One click to unsubscribe.