How to Use Gemini 4 Hydrogen at $0 Today

Yes, Gemini 4 Hydrogen is free to use. Google put gemini-4-hydrogen on the free tier of the Gemini API and in Google AI Studio on November 12, 2026, so you can call it with a free API key at $0, within rate limits. Free Gemini app users get it only through auto model selection, and Google Search uses it for everyone.

What isn’t free: Hydrogen in Google Antigravity, the new Orchestrator, Gemini in Sheets and the paid API tier. Below: each free route, its catches, and how to stretch the quota.

Is Gemini 4 Hydrogen free?

Gemini 4 Hydrogen is free on the Gemini API free tier and in Google AI Studio. Google’s pricing page lists input, output and context caching as “Free of charge” on the free tier, rate-limited. The paid tier costs $1.50 per 1M input tokens and $7.50 per 1M output tokens, flat across the 1M-token context window.

That makes Hydrogen the first Gemini 4 model with a free tier: Argon’s entry says it “is not available on the Free tier,” and our Gemini 4 Carbon free guide covers Carbon. For the full model rundown, start with what is Gemini 4 Hydrogen.

Here’s what Google published:

Every free Gemini 4 Hydrogen route

Three routes cost $0 today. Each one gives you a different slice of Hydrogen:

Route 1: The Gemini API free tier

This is the route for developers. Create a key in Google AI Studio and send requests to gemini-4-hydrogen. Google’s rate limits page lists the free tier’s qualification as an “Active project or free trial”, with verification for some projects. It doesn’t mention a payment method. Tier 1 is where billing starts.

Here’s a first call through the generateContent endpoint. Hydrogen defaults to high thinking, so set low or medium when you want fewer tokens per request:

curl -s "https://generativelanguage.googleapis.com/v1beta/models/gemini-4-hydrogen:generateContent" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [{"parts": [{"text": "Write three test cases for a token-bucket rate limiter."}]}],
    "generationConfig": {"thinkingConfig": {"thinkingLevel": "medium"}}
  }'

The SDKs use the Interactions API instead; our Gemini 4 Hydrogen API guide covers both, plus long decode and tool calls.

Route 2: Google AI Studio

AI Studio is the zero-code route: chat with Hydrogen, test each thinking level, then copy the code. The pricing page says “Google AI Studio usage is free of charge in all available regions.” AI Studio is also where you see your own rate limits, because Google doesn’t publish a fixed free-tier table for Hydrogen.

Route 3: The Gemini app, free plan

Free users can get Hydrogen in the Gemini app, but they can’t choose it. The Gemini Apps help page lists “Auto or Flash-Lite” for accounts with no Google AI plan: “Prompts that require more reasoning may use Hydrogen. If auto model selection is turned off, all responses will use Flash-Lite.” Keep auto on, and expect Hydrogen on harder prompts only.

Route 4: Google Search

Google’s Search post says Hydrogen “is now available globally in Google Search for all users.” No plan or key is needed, but you can’t control the model or build on it.

What isn’t free

Google’s launch post and the Antigravity blog put these behind a paid plan:

• Google Antigravity. Hydrogen “is rolling out across paid Google AI Pro and Ultra subscribers.” There’s no free Antigravity route to Hydrogen.
• Orchestrator. Ultra gets Orchestrator (Carbon) and Orchestrator (Argon); paid Pro gets Orchestrator (Argon). Our Gemini 4 Hydrogen Orchestrator guide explains how it hands work between Carbon and Hydrogen.
• Gemini in Google Sheets. It starts with Google AI Pro and Ultra subscribers.
• Search grounding on the API. Grounding with Google Search shows “Not available” on the free tier.
• Private data handling. On the free tier, content is “used to improve our products.” Only the paid tier says no.
• Batch API. Google lists it, at a 50% discount, under the Paid plan.

Third-party “free Hydrogen API” sites aren’t Google routes; keep your keys off them.

Gemini 4 Hydrogen free tier limits

Google hasn’t published fixed free-tier numbers for Hydrogen; limits “can be viewed in Google AI Studio.” Here’s what Google confirms:

• Limits are per project, not per API key. Extra keys in the same project don’t add quota.
• Requests per day reset at midnight Pacific time.
• Limits cover requests per minute, tokens per minute and requests per day. Cross any one and you get HTTP 429.
• Free to Tier 1 typically takes effect instantly once you link a billing account.

How to stretch the free quota

Thinking tokens count too. Five habits make the free tier last:

1. Start at medium thinking. high is the default; low uses the fewest tokens. Google suggests picking the lowest level that meets your quality bar.
2. Save xhigh for hard problems. Google warns it raises latency and token use and “can cause overthinking on simpler prompts.”
3. Reuse long prompts. Implicit caching runs on prompts of at least 4,096 tokens. Explicit caching isn’t supported on Hydrogen.
4. Route easy jobs elsewhere. Google keeps Gemini 3.5 Flash-Lite for high-volume classification and routing.
5. Back off on 429s with exponential delay, not a tight loop.

A minimal retry wrapper looks like this:

for attempt in 1 2 3 4 5; do
  code=$(curl -s -o out.json -w "%{http_code}" \
    "https://generativelanguage.googleapis.com/v1beta/models/gemini-4-hydrogen:generateContent" \
    -H "x-goog-api-key: $GEMINI_API_KEY" -H "Content-Type: application/json" \
    -d '{"contents":[{"parts":[{"text":"Summarize this changelog."}]}],
         "generationConfig":{"thinkingConfig":{"thinkingLevel":"low"}}}')
  [ "$code" != "429" ] && break
  sleep $((2 ** attempt))   # 2, 4, 8, 16, 32 seconds
done
echo "HTTP $code"

HTTP 429 Too Many Requests explains the status code. When the free tier stops being enough, our Gemini 4 Hydrogen pricing guide does the per-call math, and you can compare it with Gemini 4 Carbon pricing and Gemini 4 Argon pricing.

Moving from Gemini 3.8 Flash for free

If you’re on the free Gemini 3.8 Flash route, Google calls Hydrogen the “recommended upgrade,” and both sit on the free tier. Change the model string, drop minimal thinking (Hydrogen rejects it), remove temperature, top_p, top_k and cached_content, and raise low max_output_tokens caps to 65,536. Google reports Hydrogen is 2.5x faster than 3.8 Flash at high thinking; test your own prompts before switching.

Let your coding agent test it with Apidog CLI

Don’t spend free quota on hand-written retries. Let your coding agent set up the Hydrogen request once, assert on the response and rerun it on demand. Apidog CLI gives Claude Code, Codex or Cursor the commands, and Apidog’s Free plan covers up to 4 users.

npm install -g apidog-cli@latest
apidog login --with-token <YOUR_TOKEN>
npx -y skills add https://apidog.com   # pick your agent when prompted

Then give your agent one prompt:

In Apidog project <id>, create an environment "gemini-free" with base URL
https://generativelanguage.googleapis.com/v1beta. Add a test case for
POST /models/gemini-4-hydrogen:generateContent with the key from
{{GEMINI_API_KEY}} and thinkingLevel "medium". Assert HTTP 200, a non-empty
candidates array and a usageMetadata object. Run it with a 2-second delay
between requests and summarize any failures, including 429s.

The agent creates the environment, reads the schema, validates its payload, writes the case and runs it:

apidog environment create gemini-free --project <id> \
  --base-url https://generativelanguage.googleapis.com/v1beta
apidog cli-schema get test-case-create
apidog cli-schema validate test-case-create --file case.json
apidog test-case create --project <id> --file case.json
apidog run --test-case <caseId> --project <id> -e <envId> \
  --env-var "GEMINI_API_KEY=$GEMINI_API_KEY" --delay-request 2000 -r cli,json

--delay-request (milliseconds) spaces requests under your per-minute limit, and --env-var passes the key from your shell instead of storing it in the project. The Apidog CLI in Claude Code guide covers agent setup, and Apidog CLI in CI/CD runs the same case on every push.

FAQ

Is Gemini 4 Hydrogen free? Yes, on the Gemini API free tier and in Google AI Studio, within rate limits. Free Gemini app users get it only through auto model selection, and Google Search uses it for all users.

Do I need a credit card for the Gemini 4 Hydrogen free tier? Google’s rate limits page lists an active project or free trial as the free-tier requirement and doesn’t mention a payment method. Billing starts at Tier 1.

What are the Gemini 4 Hydrogen free tier rate limits? Google doesn’t publish a fixed table. You see your project’s limits in Google AI Studio. Limits apply per project, and daily quotas reset at midnight Pacific time.

Can I choose Gemini 4 Hydrogen in the free Gemini app? No. With no Google AI plan, you get Auto or Flash-Lite. Auto may route harder prompts to Hydrogen; with auto off, every response uses Flash-Lite.

Is Gemini 4 Hydrogen free in Google Antigravity? No. Google is rolling it out to paid Google AI Pro and Ultra subscribers, and Orchestrator is limited to those plans too.

Start free, upgrade with one string

Gemini 4 Hydrogen is the first Gemini 4 model you can build on for $0: a free AI Studio key, gemini-4-hydrogen as the model, medium thinking and a 429 backoff. For the paid side, see Gemini 4 Carbon vs Argon and what is Gemini 4 Carbon. Download Apidog and let your agent keep the Hydrogen test green as you scale.