How to Use Gemini 3.5 Flash for Free ?

Gemini 3.5 Flash launched on May 19, 2026, and Google kept the free access tier alive. Flash is callable from a free API key today; Pro arrives in June. If you want to use Flash without paying a cent, here are the five paths that actually work.

This guide focuses on practical setup for Flash specifically. By the end you’ll have a free key, working code samples, and a clear sense of where the limits hit.

Quick summary

Each path has trade-offs. The right one depends on whether you’re building, automating, or just trying Flash.

Path 1: Gemini app (the no-setup path)

If you just want to talk to Gemini 3.5 Flash, this is the fastest path.

1. Go to gemini.google.com
2. Sign in with a Google account
3. In the model selector, pick 3.5 Flash
4. Start chatting

The app handles text, image input, file uploads, and Canvas-style document editing. Free tier users get a daily message cap that resets each day. If you hit it, you wait or upgrade.

Where the app shines:

• Research and writing where you don’t need to script anything
• Image analysis without API setup
• Side-by-side comparison with other free chat models

Where it doesn’t help:

• Building software with Flash in the loop
• Batch processing
• Anything that needs a programmable response

Path 2: Google AI Studio (browser playground)

Google AI Studio is the developer playground. It’s free, web-based, and gives you raw control over every Flash parameter.

1. Open aistudio.google.com
2. Sign in
3. Click Create new prompt or pick a starter
4. In the model dropdown, choose gemini-3.5-flash
5. Type your prompt and click Run

Things you can do here that the Gemini app can’t:

• Tune temperature, top-K, top-P
• Set system instructions
• Test JSON-mode structured output
• Run multimodal prompts with multiple images
• Get the equivalent Python/Node code for your prompt
• Export the prompt to a Flash API call

AI Studio doesn’t charge for browser-based use. It’s the best free path for prompt engineering before you commit to building.

Path 3: Free Gemini 3.5 Flash API key (the build path)

This is the one most developers want. The same AI Studio account that powers the playground also issues a free API key that works against gemini-3.5-flash.

Get the key

1. In AI Studio, click Get API key in the left nav
2. Pick or create a project
3. Click Create API key
4. Copy and save it somewhere safe

That’s it. No credit card. No verification call. The key starts working immediately against Flash. Full step-by-step walkthrough in our free Gemini API key guide.

Free tier quotas for Flash

As of May 2026, Google’s free tier for gemini-3.5-flash lands at roughly:

• 1,500 requests per day
• 1M tokens per minute
• 15 requests per minute

Numbers may shift; check Google’s pricing page for the current cap before you build production code around it. For the full pricing math when you outgrow free, see our Flash pricing breakdown.

Use the Flash key

Three quick examples to confirm it works.

Python:

import os
from google import genai

client = genai.Client(api_key=os.environ["GEMINI_API_KEY"])

response = client.models.generate_content(
    model="gemini-3.5-flash",
    contents="Give me three startup ideas for API tooling in 2026."
)
print(response.text)

Node.js:

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({ apiKey: process.env.GEMINI_API_KEY });
const response = await ai.models.generateContent({
  model: "gemini-3.5-flash",
  contents: "Give me three startup ideas for API tooling in 2026.",
});
console.log(response.text);

curl:

curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-3.5-flash:generateContent" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"contents":[{"parts":[{"text":"Hello Gemini 3.5 Flash"}]}]}'

For the full setup including streaming and tool calls, see our Gemini 3.5 Flash API guide.

Test free-tier Flash requests in Apidog

If you’re going to burn 1,500 Flash requests/day, you want to make sure each one is doing useful work. Apidog gives you a workspace where you can save the Flash endpoint, store your free key as an environment variable, and replay requests without rewriting curl every time.

The flow:

1. Download Apidog
2. Create a new request, paste the curl from above
3. Move the API key into Apidog’s environment variables
4. Save the request, add response assertions
5. Run it as part of a test scenario when you change prompts

The benefit: you don’t accidentally chew through your daily Flash quota debugging the same broken request five times. Apidog caches the response history for you.

Path 4: Vertex AI new-account credits

If you create a brand-new Google Cloud account, you get $300 in credits valid for 90 days. Those credits cover Vertex AI’s hosted Gemini 3.5 Flash endpoint, which is the same model but with org-grade controls.

How to use it:

1. Sign up at cloud.google.com and confirm the $300 credit
2. Enable the Vertex AI API
3. Create a service account, give it aiplatform.user
4. Download the JSON credentials
5. Call Flash via the Vertex AI SDK

import vertexai
from vertexai.generative_models import GenerativeModel

vertexai.init(project="your-project-id", location="us-central1")
model = GenerativeModel("gemini-3.5-flash")
response = model.generate_content("Explain CAP theorem.")
print(response.text)

Vertex doesn’t have the same per-day request cap as the AI Studio free tier. You’re constrained by the dollar credit instead. At Flash’s ~$1.50/$9 per 1M tokens, $300 stretches a long way, possibly weeks of moderate use.

What to watch out for:

• The credit expires in 90 days; unused credit is forfeit
• Auto-conversion to paid happens at expiry, disable it if you don’t want surprises
• Vertex requires more setup than AI Studio, so factor that into your timeline

Path 5: Gemini CLI (the terminal path)

The Gemini CLI is Google’s open-source terminal client. It authenticates with your Google account and gives you a daily quota of 1,000 requests against Gemini’s free tier, no API key required.

Install and run:

# Install
npm install -g @google/gemini-cli

# Run
gemini

# Inside the CLI, pick gemini-3.5-flash

This is the easiest path for ad-hoc Flash scripting from your terminal. It also doubles as a free Claude Code alternative for coding work, since it can read files and run tools.

Where the CLI shines:

• Quick scripts where you’d otherwise paste into a chat UI
• Codebase grep + summarize workflows
• Daily quota separate from your API key quota

Where it doesn’t help:

• Production integrations (use the API key path instead)
• Workloads that need raw HTTP streaming control

Path 6 (bonus): OpenRouter and unlimited gateways

Several third-party gateways aggregate AI model access and route free or near-free Flash calls through their service. The pattern is covered in our Get Free Unlimited Gemini API post.

Quick caveats:

• Quality varies; some gateways throttle aggressively
• Your prompts pass through a third party, don’t send sensitive data
• Some accept payment in crypto or other workarounds

Treat this as a backup, not a primary stack. For real builds, stick with AI Studio or Vertex.

Which free Flash path should you pick?

A simple decision tree:

• Just want to chat? Gemini app
• Engineering prompts? AI Studio playground
• Building software? AI Studio API key
• Need higher quota and have a fresh GCP account? Vertex AI credits
• Scripting from your terminal? Gemini CLI
• Want raw scale and don’t mind third parties? OpenRouter and similar

Most developers end up using two together: AI Studio key for the main build, plus Vertex credits for stress testing.

Free vs paid Flash: when to upgrade

The free tier is generous. Most side projects never outgrow it. Three signals it’s time to move to paid Flash:

1. You’re hitting the 1,500 requests/day cap multiple days in a row. Upgrade. Paid Flash is cheap enough that the dev time spent dodging quotas costs more.
2. You need higher per-minute throughput. Free tier caps at 15 RPM; paid tiers go much higher.
3. You need data residency or audit logs. Move to Vertex AI on a billed account.

For paid pricing math (per-token rates, batch discount, real-world cost scenarios), see our Gemini 3.5 Flash pricing guide.

Compare the upgrade math against neighboring models too. GPT-5.5 and Claude Opus 4.7 both have their own free paths with different limits; sometimes the cheapest answer is to mix providers for different workloads. The three-way comparison breaks down where each one wins.

Tips to stretch the free Flash tier

Five small habits that make 1,500 requests/day last longer:

• Cache aggressively. Idempotent queries should hit your cache, not the API.
• Use batch mode for bulk work. The Gemini API batch mode gives you a 50% discount on non-realtime jobs; it counts against a separate quota too.
• Use structured output. A JSON schema prompt does in one call what regex parsing on free-form text does in three.
• Pre-validate inputs in Apidog. Apidog’s assertions catch malformed requests before they burn quota.
• Pick Flash over Pro by default. When Pro arrives, it’ll have tighter quotas. Stay on Flash for routine work.

FAQ

Is Gemini 3.5 Flash really free? Yes, with caps. The Gemini app, AI Studio, and the AI Studio API key all give you Flash access free with daily quotas. Paid tiers exist for higher throughput.

Do I need a credit card to get a free Flash key? No. The AI Studio free tier doesn’t require a card. Vertex AI does require a card for the $300 credit, but you won’t be charged until you upgrade.

Can I use the free Flash key in production? Technically yes, but the 1,500 requests/day cap will hit fast. For production, move to a billed account or Vertex AI.

Will Gemini 3.5 Pro be free too? Google has historically extended free-tier access to Pro models at GA. Expect Pro on AI Studio free with tighter quotas next month.

Can I use free Flash in Cursor or VS Code? Yes, both support custom API keys. The pattern is the same as our Gemini 3.0 Pro with Cursor walkthrough, just swap the model string to gemini-3.5-flash.

Are there usage logs on the free tier? Yes. Open AI Studio, click your project, and view the Activity tab.

What’s the catch? Two things. First, the free tier may use your prompts to improve Google’s models (opt-out in the AI Studio settings). Second, free-tier rate limits can shift without notice, don’t bet a launch deadline on them.

A starter project to test the free Flash tier

Pick something concrete. Build a small API that takes a URL, fetches the page, and uses Flash to summarize it.

import os, requests
from google import genai

client = genai.Client(api_key=os.environ["GEMINI_API_KEY"])

def summarize_url(url):
    html = requests.get(url).text
    response = client.models.generate_content(
        model="gemini-3.5-flash",
        contents=f"Summarize this webpage in 3 bullets:\n\n{html[:50000]}"
    )
    return response.text

print(summarize_url("https://blog.google/"))

Wire it up in Apidog to test the full request/response, then deploy it to Cloud Run or your hosting of choice. If you can ship that, you understand the free Flash path end to end.