Which Gemini Model Are You Using? How to Check, and How to Switch

Gemini does not tell you which model answered you, and on the free plan it does not let you pick one. Both of those are checkable rather than mysterious, and they explain a lot of "why was that answer bad" and "why does my picture look wrong". This page covers how to see which model you got, who is allowed to change it, and the four everyday things the choice quietly decides.

Key Takeaways

• Gemini picks the model for you by default. Google calls this auto model selection, and says prompts are "automatically routed to the most efficient model and thinking level to deliver the best possible response and preserve your usage."
• You can always see which model answered. Tap or click More under a reply, then See response details on a phone or Show thinking steps on a computer. No plan required.
• Choosing the model yourself needs a paid plan. Google's wording is flat: "If you have a Google AI plan you can manually switch models in both Chat Mode and Task Mode."
• The free plan gets auto selection or Flash-Lite, Google's fastest and least capable model. Everything above it carries a padlock.
• Turning auto selection off makes the free plan worse, not better. Google says that with it off, "all responses will use Flash-Lite."
• The model decides your image quality. Flash-Lite gives you Nano Banana 2 Lite; Flash or Pro gives you Nano Banana 2.
• The model decides whether you can make a full music track. Google says a full track needs Pro, and the free plan cannot select Pro.
• Thinking level is a second, separate dial, and it defaults to Low if you switch models without choosing one.

How to See Which Gemini Model Answered You

Open the reply and ask it. Google documents this on every platform and it needs no subscription, which makes it the one thing on this page everybody can do. The wording of the menu item differs by device, which is why people miss it:

On a computer the model name is at the end of the thinking steps rather than at the top, so scroll. On phones the details panel names the model directly. Google documents the same steps for the Gemini app and for gemini.google.com in a mobile browser.

Worth doing once before you conclude anything about Gemini's quality. If a reply came back thin and the details say Flash-Lite, you have learned something about the routing, not about the model you thought you were using.

Who Is Allowed to Change It

Only paying subscribers, and Google says so plainly rather than burying it. From its models help page, read live on November 2, 2026: "If you have a Google AI plan you can manually switch models in both Chat Mode and Task Mode." The same page's table sets out what each plan may select:

Prices from Google's subscriptions page, read the same day. There is one more condition people trip on: Google's Use Gemini Apps page says that "to switch between models in Gemini Apps, you must be signed in." Signed out, there is no picker at all.

If you do have a plan, the switch is at the bottom of the chat box. Click or tap the model name that is already showing, then pick another. Google notes that "higher-tier models marked with a lock icon require an eligible plan upgrade", so the padlock is a price tag rather than a fault. Your choice then "continues during the current conversation including if you later return to that same conversation" -- it sticks to the chat, not to your account.

What the Four Models Actually Are

Google's own one-line descriptions, in order of how hard they work. These are the words on its help page, not a benchmark:

Note what is missing: version numbers. Google deleted them from its consumer surfaces in October 2026, so "Pro" no longer tells you which Pro. For what the current top model is, see Gemini 4 Argon.

What Auto Selection Does on the Free Plan

More than you would guess, and the footnote is the part worth reading. Google's note on the free row says: "If auto model selection is turned on, Gemini will choose the appropriate model for each prompt. Most prompts will use Flash-Lite for quick answers. Prompts that require more reasoning may use Flash or Pro. If auto model selection is turned off, all responses will use Flash-Lite."

So on the free plan auto selection is the only route to a better model. You cannot ask for Flash or Pro, but Gemini can decide to use them. That makes the toggle a trap: turning auto selection off looks like taking control and actually pins you to the weakest model for everything.

Where the switch is, if you want it anyway: on a computer, Settings at the bottom left, then Usage limits, then turn auto model selection off at the bottom. The same path appears on the Android and iPhone versions of the page.

One more thing the same page says, and it only applies to subscribers: "If you have a Google AI plan and reach your usage limit, your chat will continue with Flash-Lite." Paying does not cut you off at the limit; it drops you to the model the free plan started on.

Four Everyday Things the Model Choice Decides

This is why the setting matters to someone who does not care about models at all.

Your pictures change. Google's image help page ties the image model to the chat model: Nano Banana 2 Lite "is used when the Gemini model is set to Flash-Lite", is "optimized for speed", and "isn't designed for multiple image references or for multiple edits". Nano Banana 2 arrives "when you make an image request with the Gemini model set to Flash or Pro" and brings "reliable text rendering". If your generated image mangled its text, that is the likely reason. Full detail in our Nano Banana guide.

Your music gets cut short. Google's music help page carries an "Important" note: "To create a full track, switch the model to Pro." A full track runs about two to three minutes, a short one about a minute, and the free plan cannot select Pro. Read live on November 2, 2026, Google does not publish what a free account gets instead. See Lyria 3.5, explained.

Your downloads shrink. The same image page says you can "download images at 2K resolution with a Google AI plan or at 1K without an AI plan". That one is the plan rather than the model, but it lands in the same place.

Your answers get shorter. Thinking level is a separate dial from the model: Low "provides fast, concise answers", Medium "balances processing speed and reasoning depth", High "allows the model to carefully process and reason through complex, multistep prompts". Google adds a warning and a default worth knowing: "more advanced models and higher thinking levels require more of your usage", and "if you don't select a thinking level, it will default to Low".

So Should You Pay to Choose?

Only for a reason you can name.

Stay free if your Gemini use is questions, drafts and the occasional image. Auto selection can already route you to Flash or Pro when a prompt needs it, and the thing you lose is the ability to insist. Check a reply's details once a week and you will see whether the routing is treating you well.

Pay $4.99 for AI Plus if you want a specific model on purpose. It is the cheapest tier with a picker, the cheapest route to full-length music tracks, and it also adds video generation and the Nano Banana Pro image redo. For whether free Gemini is enough in general, see is Gemini free.

Pay $19.99 for AI Pro if you want Argon, Google's top model, or model choice inside Tasks. AI Plus gets neither. For how that compares with paying OpenAI instead, see which AI chatbot you should pay for.

The bottom line: on the free plan, Gemini's model picker is a padlock and auto selection is your friend. Leave it on, learn where the response details live, and you will at least know what answered you.