Tekimax LogoSDK

Tekimax organizes AI capabilities into four distinct modalities. Each modality maps to a namespace on the client, so IDE auto-complete always shows the right methods. Using strict Capability Interfaces (VisionCapability, ImageGenerationCapability, etc.), the SDK knows exactly what your provider supports at compile time. Unsupported calls will instantly highlight as a TypeScript error!


Text (Chat)

The client.text namespace handles all Language Model interactions.

Code
import { Tekimax, OpenAIProvider } from 'tekimax-ts'; const client = new Tekimax({ provider: new OpenAIProvider({ apiKey: process.env.OPENAI_API_KEY! }) }); const response = await client.text.chat.completions.create({ model: 'gpt-4o', messages: [{ role: 'user', content: 'Explain quantum computing' }], }); // ChatResult has a flat shape — access .message directly. console.log(response.message.content);

Images

The client.images namespace unifies image generation (e.g., DALL-E) and editing.

Image Generation

Code
// Generate an image const result = await client.images.generate({ prompt: 'A futuristic city with flying cars, cyberpunk style', model: 'dall-e-3', size: '1024x1024' }); console.log(result.data[0].url);

Image Analysis (Vision)

Analyze images using multi-modal models like GPT-4o or Claude 3.5 Sonnet. The SDK normalizes the request format — OpenAI uses image_url content parts, Anthropic uses image source blocks, and Gemini uses inlineData — but you always call the same method.

Code
const analysis = await client.images.analyze({ model: 'gpt-4o', image: 'https://example.com/chart.png', // URL or Base64 prompt: 'Extract the data from this chart.' }); // ImageAnalysisResult returns .content directly (not wrapped in .message). console.log(analysis.content);

Audio

The client.audio namespace provides Text-to-Speech (TTS) and Transcription (STT) capabilities.

Text-to-Speech

Code
import { Tekimax, OpenAIProvider } from 'tekimax-ts'; const client = new Tekimax({ provider: new OpenAIProvider({ apiKey: process.env.OPENAI_API_KEY! }) }); // Convert text to speech const audio = await client.audio.speak({ model: 'tts-1', input: 'Hello, this is a generated voice.', // "alloy" is a neutral, balanced voice — good for demos. // Other options: echo, fable, onyx, nova, shimmer. voice: 'alloy', }); // Returns an ArrayBuffer — write to a file or pipe to an audio player. console.log(`Generated ${audio.buffer.byteLength} bytes of audio`);

Audio Transcription (Speech-to-Text)

Code
import fs from 'node:fs'; // Transcribe audio to text using Whisper // TranscriptionOptions.file accepts File, Blob, or Buffer. const transcription = await client.audio.transcribe({ file: fs.readFileSync('recording.mp3'), model: 'whisper-1', // 'verbose_json' returns timestamps and segment-level data, // which is useful for subtitle generation or word-level alignment. response_format: 'verbose_json', }); console.log(transcription.text); console.log(transcription.segments); // [{ start, end, text }, ...]

Video

The client.videos namespace handles video generation and analysis.

Video Generation

Code
// Generate a video from a prompt const video = await client.videos.generate({ prompt: 'A cat running in a field of sunflowers', model: 'luma-dream-machine', }); console.log(video.data[0].url);

Video Analysis (Gemini)

We support native video analysis using Gemini's multi-modal capabilities. Gemini is currently the only provider with built-in video understanding — the SDK downloads the video and converts it to inlineData automatically.

Code
import { Tekimax, GeminiProvider } from 'tekimax-ts'; const client = new Tekimax({ provider: new GeminiProvider({ apiKey: process.env.GOOGLE_API_KEY! }) }); const analysis = await client.videos.analyze({ video: 'https://cdn.example.com/beach_sunset.mp4', // gemini-1.5-flash is preferred for video analysis because it has // a 1M token context window at lower cost than gemini-1.5-pro. model: 'gemini-1.5-flash', prompt: 'Describe the scene and the lighting conditions.' }); console.log(analysis.content);

On this page