Cookie Preferences
We use cookies to personalize content and analyze traffic. See our Cookie Policy.
Gemini 3 Flash Preview (Text Generation)
Send a message to test this OpenAI-compatible chat completions model.
Complete guide to using Google Gemini 3 Flash Preview
Google's fastest multimodal AI model optimized for speed and efficiency. Process text, images, videos, and audio with low latency and cost-effective performance.
Gemini 3 Flash Preview is Google's latest lightweight multimodal AI model designed for high-speed inference and cost efficiency. Built on the Gemini architecture, Flash Preview excels at tasks requiring quick responses while maintaining strong performance across text generation, image understanding, video analysis, and audio processing. The model supports a large context window of up to 1 million tokens, making it ideal for processing lengthy documents, extended conversations, and complex multimodal inputs. The Gemini 3 Flash Preview API provides developers with OpenAI-compatible access to this powerful model, enabling seamless integration into existing applications and workflows.
Gemini 3 Flash Preview excels at natural language tasks including text generation, summarization, translation, and multi-turn conversations. With its large context window, it can maintain coherent discussions over extended interactions and process lengthy documents efficiently.
The model natively supports multiple input modalities including images, videos, and audio files. It can analyze visual content, transcribe audio, describe videos, and reason across different media types within a single conversation, making it versatile for diverse applications.
Gemini 3 Flash Preview is optimized for speed, delivering responses with minimal latency. This makes it ideal for real-time applications, chatbots, and interactive experiences where quick response times are critical for user experience.
As a lightweight model variant, Gemini 3 Flash Preview offers significantly lower per-token costs compared to larger models. This enables developers to build AI-powered applications at scale without excessive infrastructure costs.
Process entire codebases, lengthy documents, or extended conversation histories with the massive context window. The model maintains coherence and relevance across vast amounts of input data.
Send images, videos, and audio alongside text prompts without preprocessing. The model understands and reasons across multiple modalities natively, enabling rich multimodal applications.
The API follows OpenAI's chat completions format, making migration easy. Use existing SDKs and code with minimal changes to access Gemini's capabilities through apipass.dev.
Get started with Gemini 3 Flash Preview API in just a few simple steps...
Create an account on apipass.dev and obtain your API key. This key authenticates all requests to the Gemini 3 Flash Preview API and links usage to your account for billing and monitoring purposes.
Use the interactive playground to test Gemini 3 Flash Preview with different prompts, system instructions, and multimodal inputs. Experiment with temperature and other parameters to understand model behavior before integration.
Use the OpenAI Python or JavaScript SDK with apipass.dev as the base URL. The API is fully compatible with OpenAI's chat completions format, requiring minimal code changes for migration from other providers.
Deploy your integration with confidence. The API supports both streaming and non-streaming responses, handles multimodal inputs, and provides detailed usage statistics for monitoring and optimization.
As your usage grows, leverage the cost-effective pricing and high throughput of Gemini 3 Flash Preview. Monitor usage through the dashboard and optimize prompts for efficiency.
Follow these guidelines to get the best results from Gemini 3 Flash Preview API and optimize your usage for both quality and cost efficiency.
Define clear system prompts to establish the model's persona, expertise, and response format. This helps maintain consistency across conversations and improves output quality for specific use cases.
Use lower temperature (0.1-0.5) for factual, deterministic outputs like code or data extraction. Use higher temperature (0.7-1.0) for creative tasks like writing or brainstorming. Default of 1.0 provides balanced results.
For user-facing applications, enable streaming to display responses as they're generated. This significantly improves perceived responsiveness and user experience, especially for longer outputs.
Build responsive conversational AI with Gemini 3 Flash Preview's low latency and strong language understanding. The model handles multi-turn conversations naturally and can be customized with system prompts for specific personas or domains.
Process large documents, articles, or transcripts efficiently with the 1M token context window. Extract key insights, generate summaries, or answer questions about lengthy content without chunking or complex preprocessing.
Analyze visual content for descriptions, object detection, text extraction (OCR), or content moderation. The multimodal capabilities enable applications in e-commerce, media processing, and accessibility tools.
Get help with coding tasks including code generation, debugging, explanation, and refactoring. The model understands multiple programming languages and can process entire codebases within its context window.
Access Gemini 3 Flash Preview at competitive per-token rates through apipass.dev. The credit-based system provides transparent billing with no hidden costs, making it easy to budget and scale your AI applications.
Comprehensive documentation with code examples for multiple languages and frameworks. The OpenAI-compatible interface means existing tutorials and resources for chat completions APIs apply directly to apipass.dev.
Enterprise-grade infrastructure ensures high availability and consistent performance. Our support team is available around the clock to assist with integration, troubleshooting, and optimization.
All APIs require authentication via Bearer Token.
Authorization: Bearer
Generate a chat completion response using Gemini 3 Flash Preview
The API accepts a JSON payload following the OpenAI chat completions format:
1{
2 "model": "gemini-3-flash-preview",
3 "messages": [
4 {
5 "role": "system",
6 "content": "You are a helpful assistant."
7 },
8 {
9 "role": "user",
10 "content": "Hello!"
11 }
12 ],
13 "stream": false,
14 "temperature": 1,
15 "max_tokens": 4096,
16 "channel": "auto"
17}modelRequiredstringThe model ID to use for completion
"gemini-3-flash-preview"
messagesRequiredarrayArray of message objects representing the conversation history
[{"role": "user", "content": "Hello"}]streamOptionalbooleanWhether to stream the response. Defaults to false.
false
temperatureOptionalnumberSampling temperature between 0 and 2. Higher values increase randomness.
1.0
max_tokensOptionalintegerMaximum number of tokens to generate in the response.
4096
top_pOptionalnumberNucleus sampling parameter between 0 and 1.
1.0
frequency_penaltyOptionalnumberFrequency penalty between -2 and 2.
0
presence_penaltyOptionalnumberPresence penalty between -2 and 2.
0
channelOptionalstringYou may specify the corresponding provider within APIPASS via the channel parameter; these providers handle the actual image and video generation tasks. APIPASS currently offers three provider options:
The default value for the channel parameter is auto. When enabled, APIPASS automatically allocates tasks across available providers based on real-time pricing and stability metrics to balance minimal cost and reliable performance. Retain the default auto value unless you have custom routing requirements.
Available options:
auto
Each message in the messages array has the following structure:
roleRequiredstringThe role of the message author: 'system', 'user', or 'assistant'
Available options:
"user"
contentRequiredstring | arrayThe content of the message. Can be a string for text-only, or an array for multimodal content.
"What is the capital of France?"
content[].typeRequiredstringFor multimodal content: 'text' or 'image_url'
Use 'image_url' for images, videos, and audio URLs
"image_url"
content[].textOptionalstringThe text content when type is 'text'
"Describe this image"
content[].image_url.urlOptionalstringThe URL or base64 data URI of the media (image/video/audio)
Supports HTTP URLs or data:mime/type;base64,... format
"https://example.com/image.jpg"
1# Text-only request
2curl -X POST "https://api.apipass.dev/v1/chat/completions" \
3 -H "Content-Type: application/json" \
4 -H "Authorization: Bearer YOUR_API_KEY" \
5 -d '{
6 "model": "gemini-3-flash-preview",
7 "messages": [
8 {"role": "system", "content": "You are a helpful assistant."},
9 {"role": "user", "content": "What are the benefits of exercise?"}
10 ],
11 "temperature": 0.7,
12 "max_tokens": 2048
13 }'
14
15# Multimodal request with image
16curl -X POST "https://api.apipass.dev/v1/chat/completions" \
17 -H "Content-Type: application/json" \
18 -H "Authorization: Bearer YOUR_API_KEY" \
19 -d '{
20 "model": "gemini-3-flash-preview",
21 "messages": [{
22 "role": "user",
23 "content": [
24 {"type": "text", "text": "Describe this image"},
25 {"type": "image_url", "image_url": {"url": "https://example.com/image.jpg"}}
26 ]
27 }]
28 }'
29
30# Streaming request
31curl -X POST "https://api.apipass.dev/v1/chat/completions" \
32 -H "Content-Type: application/json" \
33 -H "Authorization: Bearer YOUR_API_KEY" \
34 -d '{
35 "model": "gemini-3-flash-preview",
36 "messages": [{"role": "user", "content": "Write a short story"}],
37 "stream": true
38 }'1{
2 "id": "chatcmpl-1737936000",
3 "object": "chat.completion",
4 "created": 1737936000,
5 "model": "gemini-3-flash-preview",
6 "choices": [
7 {
8 "index": 0,
9 "message": {
10 "role": "assistant",
11 "content": "Exercise offers numerous benefits for both physical and mental health...\n\n1. **Physical Health**: Improves cardiovascular health, strengthens muscles...\n2. **Mental Health**: Reduces stress, anxiety, and depression...\n3. **Weight Management**: Helps maintain healthy body weight...\n4. **Energy Levels**: Increases overall energy and reduces fatigue..."
12 },
13 "finish_reason": "stop"
14 }
15 ],
16 "usage": {
17 "prompt_tokens": 28,
18 "completion_tokens": 156,
19 "total_tokens": 184
20 }
21}idUnique identifier for the completion
objectObject type, always 'chat.completion'
createdUnix timestamp of when the completion was created
modelThe model used for the completion
choicesArray of completion choices (typically one)
choices[].messageThe generated message with role and content
choices[].finish_reasonReason generation stopped: 'stop', 'length', etc.
usageToken usage statistics for billing
xAI
xAI
Starting from
0 credits
OpenAI
OpenAI
GPT-5.6 models shared by Codex-compatible clients and direct API callers.
Starting from
0 credits
Anthropic
Anthropic
Upgrade to Fable 5.1 for sharper, more natural responses with improved context awareness and refined writing quality.
Starting from
0 credits
OpenAI
OpenAI
GPT‑6 Astra brings together years of research and big bets across pre-training, reinforcement learning, and alignment
Starting from
0 credits
Anthropic
Anthropic
Harness Fable 5 to turn raw ideas into vivid, natural-sounding text across any creative or conversational task.
Starting from
0 credits
Anthropic
Anthropic
Deploy Opus 4.8's advanced cognitive architecture to tackle your most complex reasoning and multi-step problem-solving challenges.
Starting from
0 credits
Anthropic
Anthropic
Trust Opus 5 for deep, nuanced reasoning on your hardest analytical, coding, and research problems.
Starting from
0 credits
Anthropic
Anthropic
Choose Sonnet 5 for stronger reasoning, nuanced conversation, and reliable, safety-optimized performance across everyday tasks.
Starting from
0 credits
Anthropic
Anthropic
Starting from
0 credits
OpenAI
OpenAI
Starting from
0 credits
OpenAI
OpenAI
Orchestrate autonomous AI agents with GPT-5.5 API.
Starting from
0 credits
Gemini 3 Pro Preview - Try out the latest features
Starting from
0 credits
Anthropic
Anthropic
Choose Haiku 5.5 for fast, cost-efficient responses and everyday tasks, with Claude Messages API support.
Starting from
0 credits