01
Gemini 3 Flash Preview
gemini-3-flash-preview — Google's fastest frontier model. 3x faster than 2.5 Pro, 1M context window, SWE-bench 78%, dynamic thinking levels. Input: $0.4/1M tokens, Output: $2.4/1M tokens.
gemini-3-flash-preview
Chat
Model:
Gemini 3 Series API supports chat and coding for production assistants, coding agents, support tools, and structured reasoning workflows.
Pricing details
Transparent pricing with no hidden fees. Pay as you go.
gemini-3-flash-previewInput
PoYo price
$0.400/1M tokens
1000 credits
Official price
$0.500/1M tokens
Official
You save
20%
gemini-3-flash-previewOutput
PoYo price
$2.40/1M tokens
5000 credits
Official price
$3.00/1M tokens
Official
You save
20%
gemini-3-pro-previewInput
PoYo price
$0.800/1M tokens
1000 credits
Official price
$2.00/1M tokens
Official
You save
60%
gemini-3-pro-previewOutput
PoYo price
$4.80/1M tokens
5000 credits
Official price
$12.00/1M tokens
Official
You save
60%
gemini-3.1-pro-previewInput
PoYo price
$0.800/1M tokens
1000 credits
Official price
$2.00/1M tokens
Official
You save
60%
gemini-3.1-pro-previewOutput
PoYo price
$4.80/1M tokens
5000 credits
Official price
$12.00/1M tokens
Official
You save
60%
| Model | Spec | PoYo price | Official price | You save |
|---|---|---|---|---|
| gemini-3-flash-preview | Input | $0.400/1M tokens 1000 credits /1M tokens | $0.500/1M tokens Official | 20% |
Output | $2.40/1M tokens 5000 credits /1M tokens | $3.00/1M tokens Official | 20% | |
| gemini-3-pro-preview | Input | $0.800/1M tokens 1000 credits /1M tokens | $2.00/1M tokens Official | 60% |
Output | $4.80/1M tokens 5000 credits /1M tokens | $12.00/1M tokens Official | 60% | |
| gemini-3.1-pro-preview | Input | $0.800/1M tokens 1000 credits /1M tokens | $2.00/1M tokens Official | 60% |
Output | $4.80/1M tokens 5000 credits /1M tokens | $12.00/1M tokens Official | 60% |
* Actual fees are based on the final output.
Introduction
Complete guide to using Affordable Gemini 3 Series API on PoYo
Affordable Gemini 3 Series API on PoYo
Access the full Gemini 3 Series API instantly on PoYo—no waitlist required. Choose from Gemini 3 Flash Preview, Gemini 3 Pro Preview, and Gemini 3.1 Pro Preview with 1M token context window, dynamic thinking levels, and state-of-the-art multimodal capabilities. Supports both /v1/chat/completions and Gemini Native Format endpoints for seamless integration.
Supported Gemini 3 Models on PoYo
View Documentation
02
Gemini 3 Pro Preview
gemini-3-pro-preview — Google's most advanced reasoning model. LMArena 1501 Elo, GPQA Diamond 91.9%, MoE architecture with 1T+ parameters. Input: $0.8/1M tokens, Output: $4.8/1M tokens.
03
Gemini 3.1 Pro Preview
gemini-3.1-pro-preview - Google's upgraded Gemini 3 Pro model for stronger reasoning, improved token efficiency, and more reliable agentic workflows. Input: $0.8/1M tokens, Output: $4.8/1M tokens.
Key Features of Gemini 3 Series API
Google's most capable AI models with frontier reasoning, multimodal understanding, and built-in tools.
01
1M Token Context Window
Process massive codebases, long documents, and complex multi-file projects with a 1 million token context window — the largest among frontier models. Handle entire repositories, research papers, and multi-document analysis in a single request.
- 1,000,000 token input context window
- Process entire codebases in one request
- Automatic context caching for repeated content
02
Dynamic Thinking Levels
Gemini 3 features configurable reasoning depth with the thinking_level parameter. Choose from minimal, low, medium, or high thinking levels to balance response quality, reasoning complexity, latency, and cost for each request.
- Configurable thinking depth (minimal/low/medium/high)
- Optimized cost-performance balance per request
- Deep reasoning for complex tasks, fast responses for simple ones
03
State-of-the-Art Benchmarks
Gemini 3 Pro leads the LMArena Leaderboard with a breakthrough 1501 Elo score. It achieves 91.9% on GPQA Diamond for PhD-level reasoning, 37.5% on Humanity's Last Exam, and 23.4% on MathArena Apex. Flash Preview achieves 78% on SWE-bench Verified for agentic coding.
- LMArena #1 with 1501 Elo (Pro)
- 91.9% GPQA Diamond — PhD-level reasoning (Pro)
- 78% SWE-bench Verified — agentic coding (Flash)
04
Multimodal Input Support
Process text, images, audio, video, and PDFs with Gemini 3's advanced multimodal capabilities. Use the media_resolution parameter (low/medium/high) to control vision processing and optimize token usage for your specific use case.
- Text, images, audio, video, and PDF inputs
- Configurable media resolution for cost optimization
- Advanced visual and spatial reasoning
05
Built-in Tools: Search, Code Execution & More
Gemini 3 natively supports Google Search for real-time information, Code Execution for running code, File Search for document retrieval, and URL Context for web page analysis. These built-in tools enable powerful agentic workflows without external integrations.
- Google Search — real-time web information
- Code Execution — run and test code in-context
- File Search and URL Context for document analysis
06
Dual API Endpoint Support
PoYo supports both OpenAI-compatible and Gemini Native API endpoints. Use /v1/chat/completions for drop-in OpenAI SDK compatibility, or /v1beta/models/{model} for Gemini Native Format with full feature access. Switch between endpoints without changing your credentials.
- /v1/chat/completions — OpenAI-compatible endpoint
- /v1beta/models/{model} — Gemini Native Format
- Same API key works for both endpoints
Who Can Benefit from Gemini 3 API?
01
Software Development Teams
Leverage Gemini 3 Flash Preview's 78% SWE-bench accuracy for code generation, review, and refactoring. The 1M context window handles entire repositories in a single request, making it ideal for large-scale code analysis and migration through PoYo.
02
Research & Knowledge Workers
Gemini 3 Pro's LMArena-leading reasoning combined with built-in Google Search and File Search enables powerful research workflows. Analyze complex datasets, academic papers, and multi-source information with PhD-level accuracy on PoYo.
03
AI Agent Developers
Build sophisticated agents with Gemini 3's built-in tools — Google Search, Code Execution, and Function Calling. Dynamic thinking levels let you optimize cost and latency per-request for autonomous agent workflows through PoYo.
04
Enterprise & Business Intelligence
Process massive document collections, regulatory filings, and market reports with the 1M token context window. Gemini 3's multimodal capabilities handle text, charts, PDFs, and visual data in unified analysis pipelines on PoYo.
05
Education & Content Creation
Create educational content, generate explanations, and analyze complex topics with Gemini 3's frontier reasoning. Support for audio, video, and image inputs enables rich multimodal learning applications through PoYo.
Gemini 3 Flash vs Pro, Claude Opus 4.5, GPT-5.2 — Model Comparison
How do Gemini 3 models compare to the latest frontier models? Flash Preview leads in speed-to-quality ratio, while Pro Preview tops the LMArena leaderboard with record-breaking reasoning capabilities.
| Feature | Gemini 3 Flash | Gemini 3 Pro | Claude Opus 4.5 | GPT-5.2 |
|---|---|---|---|---|
| Provider | Anthropic | OpenAI | ||
| Context Window | 1M tokens | 1M tokens | 200K tokens | 400K tokens |
| SWE-bench Verified | 78% | N/A | 80.9% | 55.6% (Pro) |
| LMArena Elo | N/A | 1501 | N/A | N/A |
| GPQA Diamond | 90.4% | 91.9% | N/A | N/A |
| Thinking Mode | Dynamic Levels | Dynamic Levels | Extended Thinking | Thinking mode |
| Built-in Tools | Search, Code, Files | Search, Code, Files | Computer Use | Function Calling |
| Multimodal | Text, Image, Audio, Video, PDF | Text, Image, Audio, Video, PDF | Text, Image | Text, Image |
| Input Price (PoYo) | $0.4 / 1M | $0.8 / 1M | $4 / 1M | $1.75 / 1M |
| Output Price (PoYo) | $2.4 / 1M | $4.8 / 1M | $20 / 1M | $14 / 1M |
| Best For | Fast coding & agents | Complex reasoning | Deep coding & agents | Knowledge work |
Benchmark data sourced from official model documentation as of January 2026. Pricing reflects PoYo API rates.
How to Use Gemini 3 Series API on PoYo
Step 1: Register & Create API Key
Sign up for a free PoYo account and generate your API key in under 2 minutes. No waitlist, no approval process required.
Create API Key →
Step 2: Top Up Credits
Add credits to your account. Gemini 3 Series uses pay-per-use pricing — Flash: $0.4/$2.4, Pro: $0.8/$4.8 per 1M input/output tokens. No subscription fees or monthly minimums.
Add Credits →
Step 3: Integrate the API
Use the OpenAI-compatible endpoint or the Gemini Native Format endpoint. Choose your model: gemini-3-flash-preview, gemini-3-pro-preview, or gemini-3.1-pro-preview. Here's a quick start example:curl --request POST \
--url https://api.poyo.ai/v1/chat/completions \
--header 'Authorization: Bearer YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data '{
"model": "gemini-3-flash-preview",
"messages": [
{"role": "system", "content": "You are a professional AI assistant."},
{"role": "user", "content": "Tell me about the latest advances in AI."}
]
}'
Gemini Native Format:curl --request POST \
--url https://api.poyo.ai/v1beta/models/gemini-3-flash-preview:generateContent \
--header 'Authorization: Bearer YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data '{
"contents": [
{"role": "user", "parts": [{"text": "Tell me about the latest advances in AI."}]}
],
"generationConfig": {
"temperature": 1.0,
"maxOutputTokens": 4096
}
}'
View Documentation →
Gemini 3 Series API Pricing: Simple and Affordable
| Model | Input | Output |
|---|---|---|
| Gemini 3 Flash Preview | $0.4 / 1M tokens | $2.4 / 1M tokens |
| Gemini 3 Pro Preview | $0.8 / 1M tokens | $4.8 / 1M tokens |
| Gemini 3.1 Pro Preview | $0.8 / 1M tokens | $4.8 / 1M tokens |
All models include:
✓ 1M token context window
✓ Dynamic thinking levels
✓ Multimodal input (text, image, audio, video, PDF)
✓ Built-in tools (Google Search, Code Execution, File Search)
✓ /v1/chat/completions and Gemini Native Format endpoints
No subscription fees — pay only for what you use. Enterprise volume discounts available for high-usage teams. Contact sales →
Frequently Asked Questions about Gemini 3 API
What is Gemini 3 Flash Preview?
Gemini 3 Flash Preview (gemini-3-flash-preview) is Google's fastest frontier model. It outperforms Gemini 2.5 Pro across many benchmarks while being 3x faster. It features a 1M token context window, dynamic thinking levels, 78% SWE-bench accuracy for coding, and multimodal support for text, images, audio, video, and PDFs.
What is Gemini 3 Pro Preview?
Gemini 3 Pro Preview (gemini-3-pro-preview) is Google's most advanced reasoning model. It leads the LMArena Leaderboard with 1501 Elo, achieves 91.9% on GPQA Diamond for PhD-level reasoning, and features a Mixture of Experts (MoE) architecture with over 1 trillion parameters. Only 15-20 billion parameters are activated per query, enabling high performance at lower computational costs.
How much does Gemini 3 API cost on PoYo?
Gemini 3 API on PoYo offers three models: Gemini 3 Flash Preview at $0.4/$2.4 per 1M input/output tokens, plus Gemini 3 Pro Preview and Gemini 3.1 Pro Preview at $0.8/$4.8 per 1M input/output tokens. No subscription fees - pay only for actual usage. Enterprise volume discounts are available for high-usage teams.
Which API endpoints are supported?
PoYo supports two endpoints for Gemini 3: /v1/chat/completions (OpenAI-compatible, works with OpenAI SDKs) and /v1beta/models/{model}:generateContent (Gemini Native Format, supports all Gemini API features including built-in tools). Both endpoints use the same API key.
What are thinking levels?
Thinking levels control the depth of Gemini 3's internal reasoning process. You can set the thinking_level parameter to minimal, low, medium, or high to balance response quality with latency and cost. Higher thinking levels produce more thorough reasoning for complex tasks, while lower levels provide faster responses for simpler queries.
Is Gemini 3 API compatible with the OpenAI SDK?
Yes. PoYo's /v1/chat/completions endpoint is fully compatible with the OpenAI SDK. Simply change the base URL to your PoYo API endpoint and use your PoYo API key. Your existing OpenAI SDK code works with Gemini 3 models with minimal changes.
Why Use PoYo for Gemini 3 API Access
01
Instant Access
Start using Gemini 3 API in under 2 minutes. No waitlist, no approval process, no enterprise contracts required. Sign up, get your API key, and start building immediately on PoYo.
02
Affordable Pricing
Access the full Gemini 3 Series starting from just $0.4/1M input tokens with Flash Preview. No subscription fees — pay only for what you use. Enterprise volume discounts available for high-usage teams.
03
Dual API Compatibility
Use /v1/chat/completions for OpenAI SDK compatibility or /v1beta/models/{model} for Gemini Native Format with full feature access. Same API key, same model — two flexible integration paths on PoYo.
04
Production-Grade Reliability
99.9% uptime SLA with automatic failover and traffic spike handling. PoYo's infrastructure is built for enterprise applications requiring consistent Gemini 3 API performance.
05
Developer Friendly
Clean REST API with comprehensive documentation, interactive playground, and code examples in cURL, Python, and JavaScript. Production-ready integration in under 30 minutes on PoYo.