Deapi
Deapi provides a unified API for running open-source AI models on distributed GPU infrastructure. This bundle gives access to text-to-image, text-to-speech, video transcription, OCR, and other AI model capabilities at lower cost than commercial providers.
Capabilities
Image generation:
- Generate images from text prompts using open-source diffusion models
- Control generation parameters (resolution, steps, guidance scale)
- Select from multiple available models
Audio:
- Convert text to speech with multiple voice options
- Transcribe audio and video files to text
Document processing:
- Extract text from images via OCR
- Process documents for structured data extraction
Model selection:
- List available models and their capabilities
- Choose models based on quality, speed, and cost tradeoffs
Workflows
Generate an image: Provide a text prompt and optionally select a model. Configure resolution and quality parameters. The response includes the generated image.
Transcribe audio: Upload an audio or video file. Select a transcription model. Receive the transcribed text with optional timestamps.
Extract text from images: Submit an image for OCR processing. Get structured text output with layout preservation where possible.
Gotchas
- Model availability: Not all models are available at all times. List available models before submitting a job.
- Processing time: Generation jobs (especially image and video) may take seconds to minutes depending on model and parameters.
- Cost varies by model: Different models have different per-request costs. Larger models and higher resolutions cost more.
- File size limits: Audio, video, and image uploads have size limits that vary by endpoint.
