ASTRAI - Advanced AI API Interface

What is ASTRAI API Interface?

The ASTRAI API Interface is a powerful, minimalist web application designed to provide seamless access to a wide array of advanced Artificial Intelligence models. It acts as a unified gateway, allowing users to interact with various AI capabilities, from deep text analysis and reasoning to image generation and speech-to-text transcription, all through a consistent and intuitive user experience.

Our interface focuses on efficiency and user control, offering features like dynamic model selection, temperature adjustments, and specialized options for different AI tasks. It's built to facilitate complex AI interactions, making advanced AI accessible for everyone.

Getting Started: How to Use ASTRAI

ASTRAI provides an intuitive interface to interact with various advanced AI models. To begin, you'll need to obtain your own API keys from the respective model publishers and an ASTRAI User ID. Follow these simple steps:

1. Obtain Your API Key(s) from Model Publishers

   ASTRAI acts as an interface, not an AI provider. This means you are responsible for acquiring API keys directly from the AI model publishers you wish to use. Each provider (e.g., OpenAI, Google, Anthropic, DeepSeek, Moonshot, xAI for Grok) has its own process for generating and managing API keys.

   • Visit the official website of your desired AI model provider.
   • Sign up for an account and navigate to their API section or developer dashboard.
   • Generate a new API key. Ensure you understand their pricing and usage policies.
   • Important: Keep your API keys secure and confidential. Do not share them publicly.

2. Get Your ASTRAI User ID

   To access the ASTRAI interface, you need a unique User ID. This ID is used to authenticate your access to our proxy server, which forwards your requests to the AI models.

   • On the initial login screen, click the "GET ID" link.
   • You will be redirected to a purchase page where you can buy a 30-day access ID via PayPal.
   • After successful payment, your unique ASTRAI User ID will be displayed on the screen and sent to your provided email address.
   • Alternatively, if you received an ID manually from an administrator, use that.

3. Enter Your ASTRAI User ID into the Login Modal

   🆔

   Upon first launching ASTRAI, or after purchasing your ID, you will be prompted to enter your User ID.

   • In the initial modal window, type your unique User ID into the designated field.
   • Click "Confirm" to proceed.
   

4. Input Your Model API Key into ASTRAI

   🔑

   Once your ASTRAI User ID is confirmed, you will see the main interface. Locate the "API Key" input field in the header/settings area.

   • Paste the API key you obtained from your chosen AI model publisher (e.g., OpenAI, Google, etc.) into this field.
   • The interface will automatically attempt to validate the key when you send your first message.
   

5. Select the Appropriate AI Model

   🧠

   Next to the API Key field, you'll find a "Model" dropdown menu.

   • Click on the dropdown and select the specific AI model you wish to interact with (e.g., gpt-4o, gemini-2.5-pro, claude-opus-4-1, dall-e-3, whisper-1).
   • The interface will dynamically adjust its input options (e.g., showing image upload for vision models, audio upload for speech-to-text models, or image generation settings) based on your selection.
   

6. Start Interacting!

   You are now ready to use the ASTRAI API Interface.

   • Type your message into the input box at the bottom of the screen.
   • If using a vision model, upload an image via the "Select Image" button.
   • If using an image generation model, describe the image you want to create. You can click the ⚙️ icon for advanced generation options.
   • If using a speech-to-text model, upload an audio file via the "Add audio" button.
   • Adjust other settings like "Temperature" or "Thinking Budget" as needed.
   • Click the "Send" button to submit your query and receive a response from the AI.

Important Notes:

• Model Compatibility: Ensure the API key you provide is valid for the selected model. An OpenAI key won't work for a Google Gemini model, and vice-versa.
• Billing: Your usage will be billed by the respective AI model provider, not by ASTRAI.

Supported Models

ASTRAI supports a diverse range of state-of-the-art AI models, categorized by their primary function:

Text Generation & Reasoning Models

• DeepSeek: deepseek-chat, deepseek-reasoner (Deep analysis and logical reasoning)
• Kimi: kimi-latest, kimi-k2-0711-preview, kimi-k2-turbo-preview, kimi-thinking-preview (Advanced conversational AI)
• OpenAI: gpt-4o, gpt-4o-mini, gpt-4.1, gpt-4.1-mini, gpt-5, gpt-5-mini, gpt-5-nano (General purpose, high-performance language models)
• Google: gemini-2.5-pro, gemini-2.5-flash, gemini-2.5-flash-lite (Powerful, multimodal models with thinking capabilities)
• Claude: claude-opus-4-1, claude-opus-4-0, claude-sonnet-4-0, claude-3-7-sonnet-latest, claude-3-5-haiku-latest, claude-3-haiku-20240307 (Leading models for complex reasoning and long contexts)
• Grok: grok-4, grok-3-mini, grok-3-mini-fast, grok-3, grok-code-fast-1 (Dynamic and engaging conversational AI, including specialized code generation)

Image & Vision Models

• Moonshot: moonshot-v1-8k-vision-preview (Image understanding and analysis)
• Grok: grok-4 (Multimodal, including vision capabilities), grok-2-image (Dedicated image generation)
• OpenAI: dall-e-3, dall-e-2, gpt-image-1 (Advanced image generation from text prompts)
• OpenAI (Vision-enabled): gpt-4.1-mini, gpt-5-nano (Text models with image input capabilities)

Audio Models

• Speech to Text: gpt-4o-mini-transcribe, gpt-4o-transcribe, whisper-1 (High-quality audio transcription)

Model Capabilities (Max Tokens)

Below is a table outlining the typical maximum input and output token limits for the supported models. Please note that these values are approximate and can vary based on specific API versions, provider updates, and the nature of the request (e.g., multimodal inputs). "Input" refers to the total context (prompt + history), and "Output" refers to the maximum length of the generated response.

Key Functions

The ASTRAI API Interface offers a rich set of functionalities:

• Text Generation & Chat: Engage in dynamic conversations, generate creative content, summarize documents, and perform complex reasoning tasks using a variety of text-based AI models.
• Image Handling (Vision): For selected models (e.g., moonshot-v1-8k-vision-preview, grok-4, gpt-4.1-mini, gpt-5-nano), you can upload images and ask the AI to analyze, describe, or answer questions about their content.
• File Upload & Analysis: Upload various document types (.doc, .docx, .js, .html, .pdf, .txt, etc.) for the AI to process, summarize, extract information, or debug code.
• Image Generation Options ⚙️: Create stunning images from text descriptions using models like dall-e-3, dall-e-2, gpt-image-1, and grok-2-image. Click the ⚙️ (gear) icon in the input area to access advanced generation options such as:
  • Number of images (1-10)
  • Image Size (e.g., 1024x1024, 1792x1024)
  • Quality (Standard, HD)
  • Style (Vivid, Natural)
  • Output Format (URL, Base64 JSON, PNG, JPEG, WebP)
  • Background (Opaque, Transparent, Auto)
  • Output Compression (0-100%)
  • Use Revised Prompt (whether to use the AI's revised prompt for generation)
  • Note: Available options depend on the selected image generation model.
• Speech-to-Text Transcription 🎵: Transcribe audio files into text using specialized models such as gpt-4o-mini-transcribe, gpt-4o-transcribe, and whisper-1, enabling voice-to-text applications. Simply upload your audio file and optionally provide a prompt.
• Copy Code 📋: Easily copy code blocks generated by the AI to your clipboard with a single click. A "Copy" button appears on hover over any code block.
• Run HTML Code 💻: For HTML code snippets generated by the AI, a "RUN HTML" button will appear. Click it to instantly preview the HTML output in an isolated iframe, allowing for quick testing and visualization.
• Copy & Regenerate 📋🔄: After an AI response, you'll find action buttons:
  • 📋 Copy: Copies the AI's response text to your clipboard.
  • 🔄 Regenerate: Re-sends the previous user message to the AI to generate a new response. This is useful for getting alternative answers or refining the output.
• Reasoning & Thinking Budget: For advanced models (e.g., Gemini 2.5, Claude Opus/Sonnet, Grok-3 series), control the AI's "thinking budget" to influence the depth and complexity of its reasoning process, with options to view the internal thought process.
• Chat Management: Create, load, edit, import, and export chat sessions for seamless workflow and data portability.
• Customizable Settings: Adjust model temperature, select different AI models, and switch between various UI themes to personalize your experience.

Chat History Storage Notice

Your chat history is stored locally in your browser.

This means:

• Your conversations remain private and are not uploaded to our servers.
• The history will be permanently deleted if you clear your browser's data (e.g., cookies, site data, local storage).
• If you use multiple browsers or devices, your chat history will be unique to each one.

Understanding API Keys

API Keys are unique identifiers that authenticate your requests to the AI services. They act as a password, granting you access to the AI models and ensuring that your usage is tracked and billed correctly by the respective AI providers (e.g., OpenAI, Google, Anthropic, DeepSeek, Moonshot, Grok).

Key Characteristics:

• Authentication: Your API Key verifies your identity to the AI service. Without it, requests will be rejected.
• Usage Tracking: All API calls made with your key are associated with your account, allowing for accurate billing and usage monitoring.
• Confidentiality: API Keys should be treated with the same level of security as passwords. They should never be shared publicly or exposed in client-side code where they can be easily accessed.
• Provider-Specific: Each AI provider (OpenAI, Google, etc.) issues its own API Keys. You will need a valid key for the specific model you wish to use.

In the ASTRAI interface, you enter your API Key directly. This key is then used to authenticate your requests to the backend proxy, which forwards them to the respective AI service.