Yes, Gemini 4 Hydrogen is free — in three places, with real limits. Google launched it on 12 November 2026, and it is the first Gemini 4 model free users can actually use: a rate-limited free tier in Google AI Studio and the Gemini API, automatic routing in the free Gemini app, and free access in Google Search for everyone. The catch is the size of the free allowance: 20 API requests a day, and in the app you cannot pick the model yourself. This article covers exactly what is free, the numbers, and what you cannot do without paying.
What Is Gemini 4 Hydrogen?
Gemini 4 Hydrogen is Google’s fast execution model in the Gemini 4 family. While Carbon and Argon handle deep, long-horizon reasoning, Hydrogen is built for rapid execution across everyday coding and agentic workflows: repository coding and debugging, interactive web development, and multimodal work.
The headline numbers, all from Google’s launch post on 12 November 2026:
- 1 million-token input context, with up to 1 million output tokens.
- $1.50 per million input tokens and $7.50 per million output tokens, flat across the whole context window.
- 51 on the Artificial Analysis Intelligence Index at its default high thinking level — one point above Argon (50), and 54 at the new xhigh thinking level.
- 2.5x faster than Gemini 3.8 Flash, with Terminal-Bench 4.1 jumping from 18.2% to 57%, and fabricated claims cut by more than 5x.
- Four thinking levels: low, medium, high (default), and xhigh — so you pay only for the reasoning depth a job needs.
These are Google’s own figures, not independently reproduced tests. The useful fact for most readers is simpler: Hydrogen is the upgrade path from Gemini 3.8 Flash, it costs twice what Flash costs today, and it is the only Gemini 4 model with a free tier.
Where Gemini 4 Hydrogen Is Free
1. Google AI Studio (free, no card)
AI Studio is the fastest way to try it. Open Google AI Studio, sign in with a Google account, and pick gemini-4-hydrogen in the model dropdown. No billing account is needed for the free tier. Standard input and output tokens cost $0, subject to the rate limits below.
2. Gemini API (free tier, rate-limited)
The same free tier works through the API. Create a key in AI Studio, install the SDK (pip install -U google-genai), and call gemini-4-hydrogen. You do not need to link a Cloud Billing account. The limits are the real constraint — see the table below.
3. Gemini app (free, Auto only)
Free users in the Gemini app get Hydrogen through Auto — the app’s automatic model selection. When a prompt needs more reasoning, Auto may route it to Hydrogen. Most free-app prompts still go to Flash-Lite, and turning Auto off routes everything to Flash-Lite. You cannot pick Hydrogen yourself on the free plan. The model picker arrives with Google AI Plus at $4.99 a month.
4. Google Search (free, for everyone)
Hydrogen now powers Google Search for all users globally — Google calls it the biggest model upgrade to Search yet. You do not need to do anything: ask a question in Search and the improved model answers it, with more visual responses and a wider range of links.
Is Gemini 4 Hydrogen free? Yes — in Google AI Studio and the Gemini API on a rate-limited free tier (20 requests a day), in the free Gemini app when Auto picks it, and in Google Search for all users. It is not unlimited free: the API free tier caps at 20 requests a day, and free app users cannot choose the model.
Free-Tier Limits (the Numbers That Matter)
| Limit | Gemini 4 Hydrogen (free tier) |
|---|---|
| Requests per minute (RPM) | 5 |
| Input tokens per minute (TPM) | 250,000 |
| Requests per day (RPD) | 20 |
| Price (standard tokens) | $0 |
| Billing account needed | No |
Three things to know about these limits:
- Quotas are per project and per model. Creating another API key in the same Google Cloud project does not double the allowance.
- The daily count resets at midnight Pacific time. Hitting any limit returns a 429 RESOURCE_EXHAUSTED error — wait for the reset, or switch to a model with quota left (Flash-Lite allows 500 requests a day).
- Batch and Flex requests need paid access, and Search/Maps grounding is unavailable on the free tier.
One more condition: content sent through the unpaid API service may be used by Google to improve its products. Paid-tier content is not used this way. Do not send private or sensitive material through the free tier.
What It Costs After Free
| Model | Input / 1M tokens | Output / 1M tokens | Free API tier |
|---|---|---|---|
| Gemini 4 Hydrogen | $1.50 | $7.50 | Yes |
| Gemini 4 Argon | $2 ($4 from 1 Dec) | $10 ($20 from 1 Dec) | No |
| Gemini 4 Carbon | $2 ($4 from 1 Dec) | $10 ($20 from 1 Dec) | No |
| Gemini 3.8 Flash | $0.75 ($1.50 from 1 Jan 2027) | $3.75 ($7.50 from 1 Jan 2027) | Yes |
Cached input costs $0.075 per million tokens through implicit caching. Argon and Carbon’s introductory prices end on 30 November 2026 — from 1 December they double. Flash rises to Hydrogen’s price on 1 January 2027, which is when Hydrogen becomes the obvious default for anyone still on Flash.
Note: a Google AI Pro or Ultra subscription does not raise your API free tier. App subscriptions and API billing are separate — API limits depend only on your Cloud project’s billing tier.
Gemini 4 Hydrogen vs Argon vs Carbon
| Hydrogen | Argon | Carbon | |
|---|---|---|---|
| Role | Fast everyday execution | Deep reasoning | Hardest problems |
| Price (input/output per 1M) | $1.50 / $7.50 | $2 / $10 (→ $4 / $20) | $2 / $10 (→ $4 / $20) |
| Free API tier | Yes (20 req/day) | No | No |
| Free app access | Via Auto | No (Pro plan) | No (Ultra plan) |
| AA Intelligence Index | 51 (54 at xhigh) | 50 | 60 |
| Context window | 1M tokens | 1M tokens | 1M tokens |
| Best for | Coding, agents, daily work | Multi-step reasoning | Long-horizon problems |
Hydrogen is the only one of the three a free user can touch. Argon and Carbon need a paid plan or paid API access.
Honest Caveats
- The benchmarks are Google’s own. The AA Index, Terminal-Bench and HLE-Diamond figures come from Google’s launch post and its chosen evaluations. They have not been independently reproduced. Treat them as direction, not proof.
- The free app is mostly Flash-Lite. Auto routes most free-app prompts to Flash-Lite; Hydrogen appears only when the prompt needs more reasoning. If you want Hydrogen on demand in the app, that is what the $4.99 Plus plan is for.
- 20 requests a day is small. It is enough to test the model and run small scripts, not to build a product on. For repeated short prompts, Flash-Lite’s 500 free requests a day goes further.
- India availability is not separately confirmed. Google says the rollout is global, but no source we checked confirms India rollout timing or India pricing. The global-rollout wording is Google’s, not a checked India fact.
- Free-tier data may train Google’s products. That is the trade for $0 access. Keep sensitive prompts out of the free tier.
FAQ
Yes, with limits. It is free in Google AI Studio and the Gemini API on a rate-limited free tier (5 requests a minute, 20 a day), in the free Gemini app when Auto selects it, and in Google Search for all users. It is not unlimited.
Open Google AI Studio, sign in with a Google account, create an API key, and select gemini-4-hydrogen. No billing account or card is needed for the free tier.
5 requests per minute, 250,000 input tokens per minute, and 20 requests per day, per project. The daily count resets at midnight Pacific time.
No. Free app users get Hydrogen only through Auto, the automatic model choice. The model picker needs Google AI Plus ($4.99 a month).
No. Argon and Carbon need paid API access or a Google AI Pro (Argon) or Ultra (Carbon) plan. Hydrogen is the only Gemini 4 model with a free tier.
$1.50 per million input tokens and $7.50 per million output tokens, flat across the 1M-token context. Cached input is $0.075 per million.
Gemini 4 Hydrogen is the first Gemini 4 model worth trying without paying — and the free tier is genuinely usable, not a demo. Start in AI Studio: 20 requests a day is enough to feel the speed and the 1M context on real code and documents. Use Google Search and the Gemini app for everyday questions — both are free and need no setup. Pay only when the 20-a-day cap starts getting in your way; at $1.50/$7.50 it is still the cheapest Gemini 4 model, and from 1 January 2027 it costs the same as the Flash model it replaces.