Skip to content
Chat model

Gemini 3 Pro

Chat model

Available on Ofrai — chat, images, and video in one place with shared credits.

Open chat studio

Chat and video use credits from your Ofrai balance.

Input/Output 1.46278/8.7767

About 1.46278 credits per short prompt and 8.7767 credits per short reply (per ~1,000 characters of text).

Commercial use depends on the selected model and model terms.

Gemini 3 Pro example on Ofrai

About Gemini 3 Pro on Ofrai

Gemini 3 Pro on Ofrai is Google's flagship reasoning model of the 3 generation, the strongest tier in that generation and the one to reach for when a wrong answer costs more than a slower reply. Where Gemini 3 Flash handles everyday drafting and question answering, 3 Pro is built for multi-step analysis, careful instruction following and problems where one early mistake quietly poisons everything after it. Every request is billed from the shared Ofrai credit balance, so escalating to it stays on one ledger.

About 1.46278 credits per short prompt and 8.7767 credits per short reply (per ~1,000 characters of text).

Gemini 3 Pro model facts

Model ID
gemini-3-pro
Modality
Chat model
Released
Not published
Ofrai pricing
About 1.46278 credits per short prompt and 8.7767 credits per short reply (per ~1,000 characters of text).
Web search
Not available

Best for

  • Multi-step analysis where an early error would silently corrupt the conclusion
  • Complex instructions carrying many constraints that must all hold at once
  • Code, maths and logic problems where the working matters as much as the answer
  • High-stakes drafting such as contracts, proposals and analysis memos

Not the best fit when

  • High-volume classification or routing, where a Lite tier costs far less
  • Routine everyday Q&A and first-draft work a Flash model already handles well
  • Latency-sensitive interactive flows where a heavier model's extra wait is felt

How it differs

Compared with Gemini 3 Flash, it is the same generation one tier higher: Flash is the balanced default for everyday work, and 3 Pro is where you go when the reasoning chain rather than the wording is what the task turns on.

Example prompts

  • You are a [role]. Work through this in numbered steps and show your reasoning at each one, then give a final answer. Mark every step where you had to assume something.
    
    [TASK]
  • Below are [N] constraints that must all hold. Produce an output that satisfies every one, then list each constraint next to the part of your output that satisfies it.
    
    [CONSTRAINTS]
  • Find the flaw in the argument below. Do not rewrite it — name the specific step where it breaks, explain why that step is wrong, and give the smallest correction that repairs it.
    
    [ARGUMENT]
  • Review the draft below as a sceptical reviewer. List errors of fact, unsupported claims and unclear reasoning as separate sections, then give a verdict of ship, revise or reject with one sentence of justification.
    
    [DRAFT]

FAQ

What is Gemini 3 Pro used for?
Gemini 3 Pro is Google's top reasoning tier for the 3 generation, used when the cost of a wrong answer outweighs the cost of waiting. Typical jobs are multi-step analysis, code and logic problems, long chains of dependent instructions and high-stakes drafting. It is slower and dearer per call than Gemini 3 Flash, so it is usually reserved for the requests that genuinely need it rather than used as a blanket default.
How does Gemini 3 Pro differ from Gemini 3 Flash?
Both sit in the 3 generation and occupy the same role — general-purpose work — at different quality levels. Flash is the balanced tier: fast, cheaper and good enough for most drafting and Q&A. Pro spends more effort per request and holds up better on long dependent chains, dense constraints and problems where an early mistake would propagate. Choose Flash by default and escalate the hard cases.
Should I use Gemini 3 Pro for everything?
No. Running every prompt on Pro wastes budget on work a Flash model already does well, and it slows down interactive work for no benefit. A common pattern on Ofrai is to default to Gemini 3 Flash and escalate only the prompts that fail, plus routing known-hard categories straight to Pro. That keeps the average cost down while reserving the expensive model for where it changes the outcome.
Is Gemini 3 Pro the newest model on Ofrai?
No. It belongs to the 3 generation, and this page also covers the 3.1, 3.5, 3.6, 3.7 and 3.8 generations. Newer does not automatically mean better for your task: the right pick depends on whether you want the newest capability, the Pro reasoning tier, or a cheaper Flash tier for volume. Compare newer entries against your own prompts rather than trusting the version number.
How are Ofrai credits charged for Gemini 3 Pro?
Ofrai runs one shared credit balance across all models, so there is no separate Gemini wallet. Credits are drawn at the platform's published rate for that model when the request is sent, and the same balance also funds image, video and audio work. Because Pro calls cost more per request than Flash or Lite, keeping Pro usage targeted is the most effective way to control spend.
Can Gemini 3 Pro handle images and documents?
This listing is registered as a text chat model on Ofrai, so prompts are text. Google's Gemini family is itself broadly multimodal, but what a given platform exposes depends on its own integration. Treat multimodal support as something to verify for the deployment you actually use rather than assuming it. Upload controls, if present, are surfaced separately from the model entry itself.

Related models

More models in the same modality on Ofrai.

Compare

Related guides

Use cases

Tools