Gemini 3 Pro Preview
Gemini 3 Pro is Google DeepMind's flagship model with state-of-the-art multimodal reasoning across text, images, video, audio, and code. Access it through Velokey's unified endpoint with below-list pricing.
Use cases
Learning & multimodal analysis · Building & vibe coding
Input
Text, image
Output
Text
Billing
Per token
Start a conversation
Type a message below to begin
Sign up to get $0.5 in free credits — about 20 free images
Pricing details
Transparent usage-based pricing with no hidden fees.
Detailed Pricing
Usage-based pricing breakdown
| Spec | Price | Official | Savings |
|---|---|---|---|
| Input/1M tokens | 1.6 Credits≈$1.6 | 2 Credits | -20% |
| Output/1M tokens | 9.6 Credits≈$9.6 | 12 Credits | -20% |
| Cache Read/1M tokens | 0.16 Credits≈$0.16 | 0.2 Credits | -20% |
Billing Rules
- Output price varies by model, resolution, quality, duration, or token usage.
- Uploaded reference files may be charged separately when supported.
- Add-on capabilities such as web search or image search may be billed per request.
- Token-based models are billed by input and output tokens.
- If the primary route is unavailable, Velokey may automatically switch to a stable fallback route when possible.
- Usage is pay-as-you-go with transparent credit deduction.
* Final cost depends on the generated result.
Related models
Explore other models in the same family.
GoogleSave 20%Gemini 3.1 Pro Preview
Gemini 3.1 Pro delivers reasoning with unprecedented depth, top-tier agentic coding, and multimodal understanding of text, image, audio, video and PDF over a 1M-token context.
- GoogleSave 20%
Gemini 3.1 Flash Lite Preview
Gemini 3.1 Flash is Google's most cost-efficient model in the 3 series, with flexible reasoning, multimodal input and a 1M-token context for high-volume, low-latency tasks.
- GoogleSave 20%
Gemini 3 Flash Preview
Gemini 3 Flash brings Pro-grade reasoning at Flash-level latency and cost — strong coding (SWE-bench 78%), 3x faster than 2.5 Pro, and multimodal understanding.
- GoogleSave 20%
Gemini 3.5 Flash
Google's high-performance dialogue model is fast and highly accurate, ideal for text generation, Q&A, and multi-turn conversations.
Affordable Gemini 3 Pro API
Access Google DeepMind's Gemini 3 Pro on Velokey — state-of-the-art multimodal reasoning, advanced agentic coding, a 1M-token context, and a dedicated Deep Think mode, with a unified endpoint and below-list pricing for production apps.
Meet Gemini 3 Pro and its API access
Gemini 3 Pro is Google DeepMind's flagship model released on November 18, 2025, delivering state-of-the-art multimodal reasoning across text, images, video, audio, and code. It significantly outperforms Gemini 2.5 Pro on every major benchmark — topping LMArena at 1501 Elo, scoring 91.9% on GPQA Diamond and 76.2% on SWE-bench Verified — and ships with a 1M-token context window, advanced agentic and coding abilities, enhanced tool use, and improved factual accuracy with reduced hallucinations. A dedicated Deep Think mode pushes reasoning even further. Velokey provides an affordable, easy-to-integrate Gemini 3 Pro API: a unified chat endpoint, a clear request flow, and below-list pricing for production systems, agent pipelines, and automation tools.
State-of-the-art multimodal reasoning
Reasons across text, images, video, audio, and code in a single flow. Tops LMArena at 1501 Elo, with 91.9% GPQA Diamond, 81% MMMU-Pro, and 87.6% Video-MMMU — a clear leap over Gemini 2.5 Pro on every major benchmark.
View Documentation →
Agentic coding & long-horizon planning
Advanced agentic and coding abilities with enhanced tool use and planning over longer horizons — 76.2% on SWE-bench Verified and 1487 Elo on WebDev Arena, and it pairs with the Google Antigravity agentic dev platform.
View Documentation →
1M-token context understanding
A 1M-token context window with improved factual accuracy and reduced hallucinations, reasoning coherently across entire codebases, long videos, and large document sets without losing the thread.
View Documentation →
Deep Think mode for the hardest problems
An enhanced reasoning mode that lifts Humanity's Last Exam to 41.0%, GPQA Diamond to 93.8%, and reaches 45.1% on ARC-AGI-2 with code execution — for complex, high-stakes reasoning that rewards deeper deliberation.
View Documentation →
How to deploy the Gemini 3 Pro API on Velokey
Get started with just a few steps.
Sign up and get an API Key
Sign in to the Velokey console to generate an API Key that authenticates every request sent to Gemini 3 Pro.
Configure request parameters
Set model to gemini-3-pro and submit messages with text, images, video, audio, or code. Use the 1M-token context for very long inputs and enable Deep Think mode for the hardest reasoning.
Integrate and start calling
Send requests to Velokey's unified /v1/chat/completions endpoint to power assistants, agents, and automation, with usage managed in one console.
Real-world use cases for Gemini 3 Pro
From learning to building and planning, covering many high-autonomy scenarios.
Learning & multimodal analysis
Multimodal document analysis, interactive visualizations, and translation — turn dense text, images, and video into clear, interactive learning.
Building web UIs & interactive apps
Web UI generation, vibe coding, and interactive apps — pairs with the Google Antigravity agentic dev platform to go from idea to working app.
Workflow automation & planning
Multi-step workflow automation and task management with enhanced planning and tool use for reliable execution over longer horizons.
Hard reasoning with Deep Think
Deep Think mode for the hardest science, math, and analysis problems — 41.0% on Humanity's Last Exam and 93.8% on GPQA Diamond.
Video & image understanding
Strong video and image comprehension — 87.6% on Video-MMMU and 81% on MMMU-Pro — fit for media analysis and visual QA.
Agentic coding & engineering
76.2% on SWE-bench Verified with advanced agentic coding — fit for repository-scale development, refactoring, and bug fixing.
Why choose Velokey for Gemini 3 Pro
Discount on list price
Call Gemini 3 Pro at a lower price on Velokey — pay-as-you-go with no forced subscription, reducing upfront cost.
Unified API interface
One API Key accesses Gemini 3 Pro and other models — no juggling multiple accounts and keys, simplifying integration.
Complete docs and migration guidance
Endpoint references, parameter details, and migration examples help you adopt multimodal inputs, long context, and Deep Think smoothly.
Stable and highly available
Automatic failover and stable fallback routes maintain availability when the primary route is down, keeping production reliable.
API Reference
Complete API call examples and parameter descriptions
Endpoint
https://api.velokey.ai/v1beta/models/{model}:generateContentAuthentication
Bearer YOUR_API_KEYRequest Example
curl https://api.velokey.ai/v1beta/models/{model}:generateContent \
-X POST \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gemini-3-pro-preview",
"messages": [{"role": "user", "content": "Hello!"}],
"temperature": 0.7
}'Response Example
{
"id": "req_abc123",
"model": "gemini-3-pro-preview",
"created": 1234567890,
"data": { ... }
}Gemini 3 Pro API FAQ
What's improved over Gemini 2.5 Pro?
Gemini 3 Pro significantly outperforms Gemini 2.5 Pro on every major benchmark. It tops LMArena at 1501 Elo, scores 37.5% on Humanity's Last Exam (no tools), 91.9% on GPQA Diamond, 23.4% on MathArena Apex (SOTA), 81% on MMMU-Pro, 87.6% on Video-MMMU, 72.1% on SimpleQA Verified, 76.2% on SWE-bench Verified, and 1487 Elo on WebDev Arena — with stronger agentic and coding abilities and improved factual accuracy.
What is Deep Think mode?
Deep Think is an enhanced reasoning mode for the hardest problems. It lifts Humanity's Last Exam to 41.0%, GPQA Diamond to 93.8%, and reaches 45.1% on ARC-AGI-2 with code execution — best suited to complex, high-stakes reasoning where extra deliberation pays off.
How does the 1M-token context work?
Gemini 3 Pro ships with a 1M-token context window to reason coherently across entire codebases, long videos, and large document sets, with improved factual accuracy and reduced hallucinations over longer inputs.
What is it good at?
State-of-the-art multimodal reasoning across text, images, video, audio, and code, plus advanced agentic and coding abilities with enhanced tool use and long-horizon planning — fit for learning, building interactive apps, and multi-step workflow automation, and it pairs with the Google Antigravity agentic dev platform.
How is it billed on Velokey?
This page's pricing area shows the current Velokey price, billed by input and output tokens with a discount on list price. You pay as you go with one API Key, no forced subscription.
Start building with Gemini 3 Pro Preview today
Use one API key to access low-cost, stable AI models through Velokey.