Google uses the Gemini name across a consumer app, paid Google AI plans, Workspace features, and developer APIs. Those surfaces do not always expose the same models, limits, or data controls.
This guide was checked on July 31, 2026. It answers the practical questions first and links to the live Google source for details that can change.
Gemini FAQ: quick answers
| Question | Short answer |
|---|---|
| Which models appear in Gemini Apps? | Google’s current help page lists Gemini Flash-Lite, Flash, and Pro. |
| Do model names and limits stay fixed? | No. Google explicitly says names, versions, availability, and limits may change. |
| Which consumer plans exist? | Google currently describes no-AI-plan access plus Google AI Plus, Pro, and Ultra. Availability and benefits vary. |
| Is Gemini Apps the same as the Gemini API? | No. The app uses consumer-facing model choices; the API exposes versioned developer models with separate pricing and lifecycle rules. |
| Can Google retain Gemini chats? | Retention depends on the surface and settings. Some human-reviewed data may remain after activity is deleted. |
| Are work and school accounts handled like personal accounts? | Not necessarily. Workspace agreements and administrator controls can create different data-handling terms. |
Which Gemini model should I choose in the app?
Google’s Gemini Apps limits and upgrades page currently describes three app-level choices:
- Flash-Lite for fast, lightweight everyday work.
- Flash for a balance of speed and reasoning.
- Pro for harder math, coding, file, image, and video tasks where a slower answer may be worthwhile.
Deep Think is a reasoning mode rather than a separate everyday model choice. Google’s help page currently places it behind AI Ultra and requires the Pro model.
The safest way to choose is to match the live model label to the task, then confirm the current limit shown in your account. Do not rely on a screenshot or an old article for a fixed prompt quota.
Is Gemini 3.1 Pro the latest Gemini model?
The answer depends on the surface.
Gemini Apps uses simplified labels such as Flash and Pro. Google’s developer documentation uses versioned model IDs and separately marks preview, stable, deprecated, and shut-down models. A model described as current in the app is not automatically the correct API model ID.
For application development, use the live Gemini API models directory and the latest-model guidance before choosing or migrating a model. For the consumer app, use the model menu and the Gemini Apps help page.
What is the difference between Gemini Apps and the Gemini API?
| Surface | Designed for | Where to verify current facts |
|---|---|---|
| Gemini Apps | Personal chat, files, research, creative work, and connected-app tasks | Gemini Apps Help and the model menu in your account |
| Gemini in Workspace | Work or school workflows governed by an organization’s edition and admin controls | Google Workspace documentation and your administrator |
| Gemini API | Software integrations using versioned model IDs | Google AI for Developers |
| Vertex AI | Google Cloud deployment and governance workflows | Vertex AI documentation and Cloud pricing |
Treat access, billing, context windows, and data handling as surface-specific. A consumer subscription does not automatically include API usage, and an API model ID does not prove that the same label appears in Gemini Apps.
How much does Gemini cost?
Google currently groups personal access into:
- access without a Google AI plan;
- Google AI Plus;
- Google AI Pro;
- Google AI Ultra.
Higher plans provide expanded limits and different feature bundles, but prices, storage, regional availability, and benefits can change. Check the live Google AI plans page for your region before subscribing.
API usage is billed separately. Check Gemini API pricing for the exact model ID, input type, context tier, and output type you plan to use.
Does the free Gemini app include Pro?
Google’s current help table shows access to Flash-Lite, Flash, and Pro across the no-plan and paid-plan rows, while paid plans increase usage and unlock or expand some features. That does not mean every account receives the same quota or feature availability.
Google says access may change because of testing, experimentation, availability, and compute-based limits. Verify the live model menu and limit message in your own account.
What context window does Gemini Apps provide?
Google’s current help page lists different file-reading context windows by plan:
- 32K tokens without a Google AI plan;
- 128K tokens with AI Plus;
- 1 million tokens with AI Pro or AI Ultra.
These figures describe the current Gemini Apps plan table, not every Gemini API model. Recheck the official page before using them for procurement or workflow design.
Does Google use Gemini conversations to improve its services?
Google’s Gemini Apps Privacy Hub says data handling depends on the feature, account, settings, and whether human review is involved.
For a personal account:
- Review Gemini Apps Activity and the available auto-delete controls.
- Avoid entering confidential information you would not want a reviewer to see.
- Check connected-app permissions before asking Gemini to use data from another service.
- Use a temporary chat or turn off activity where appropriate, while understanding that limited retention may still apply for service and safety purposes.
Deleting activity is not a promise that every related copy disappears immediately. Google’s privacy page says chats reviewed by humans and related data may be retained for up to three years.
Can I delete Gemini chat history?
You can delete chats and manage auto-delete settings in Gemini Apps Activity. Google’s current Privacy Hub says the default auto-delete period can be changed and chats can be deleted manually.
This control applies to the activity associated with the covered Gemini Apps. Data handled by a connected app, a Workspace organization, or another Google service may follow a different policy.
What changes for Google Workspace accounts?
Work and school accounts can have different availability, retention, and administrator controls from personal accounts. Do not use the consumer Privacy Hub alone to make a Workspace compliance decision.
For organization-sensitive work:
- confirm the Workspace edition and enabled Gemini features;
- review the organization’s agreement and data region requirements;
- ask the administrator which connected apps and retention controls are enabled;
- keep regulated or confidential data out of prompts until those controls are verified.
Can Gemini use Gmail, Docs, and other connected apps?
Eligible users can connect supported Google apps and, in some cases, third-party services. The Gemini Apps Privacy Hub explains that Gemini and the connected service each process shared data under their applicable policies.
Before connecting an app, check:
- what data Gemini can retrieve or act on;
- whether the account is personal, work, or school managed;
- which activity and personalization settings are enabled;
- whether the task can be completed with a smaller data scope.
Does Gemini generate images and video?
Google offers generative-media models and features across Gemini, Google Vids, Flow, and developer products. The exact model, plan, region, and eligibility boundary varies by surface.
For developer use, the Gemini API models directory currently separates image and video models from the general text-and-reasoning family. For consumer use, check the feature shown in your account instead of assuming that an API model is included in your plan.
If your task is specifically video production, use the AI video workflow guide to compare model access, iteration cost, and review steps. For a model-focused route, see the Gemini 3.1 guide.
Can I use Gemini through Flowith?
Model availability inside Flowith can change independently from Google’s consumer app and API catalogs. Check the live Flowith model selector for the model available to your account and session. A Google announcement or developer model ID should not be treated as proof of Flowith access.
Use Flowith’s canvas when the value comes from organizing prompts, sources, branches, and outputs in one workspace. Use Google’s live documentation when the decision depends on Gemini pricing, quotas, model lifecycle, or privacy terms.
Source checklist
These first-party pages support the time-sensitive answers in this guide:
- Gemini Apps limits and upgrades
- Gemini Apps Privacy Hub
- Google AI plans
- Gemini API models
- Gemini API latest-model guidance
- Gemini API pricing
Recheck the relevant source before making a billing, privacy, or production migration decision.