[Feature] Add Straico as AI Provider

[luisblop]

What do you need?

Straico native AI Provider Integration in Deepchat

Platform

Win/macOS/Linux

Is your feature request related to an existing issue?

No, it is a new feature request

Useful resources

Main Page: https://www.straico.com/api
Discord server: https://discord.gg/straico-community-1118690015443161130
API Documentation: https://documenter.getpostman.com/view/5900072/2s9YyzddrR
Obtain API key: https://platform.straico.com/settings-api

• API Base URLs:
  • https://api.straico.com/v1
  • https://api.straico.com/v2 (for OpenAI-compatible endpoints)

Straico's Model Selection Guide: https://straico.com/multimodel
📨 Contact Straico Team: product@straico.com

Why we need it

Straico offers a unified API platform that aggregates multiple AI models and services, providing Deepchat users and Straico community with these key capabilities :

• Chat Completions: Advanced conversational AI with multiple model options
• Model Discovery: Access to multiple AI models through a single API
• Multi-modal capabilities like web search, reasoning, vision, deep research and tool calling
• Image Generation: High-quality image creation from image and text
• TTS via ElevenLabs voices
• Video Generation: Image-to-video synthesis with customizable parameters
• Straico RAG: Create new and chat with RAG systems created on the Straico platform
• Straico Agents: Communicate with agents created on the Straico platform

Desired Solution

Before : No Straico AI provider integration
After : Adds Straico AI as a native AI provider with comprehensive multi-modal AI capabilities

Implementation Strategy:

• Used direct API integration for better control over model selection
• Utilize Straico's official API endpoints, including the v2 family of OpenAI-Compatible endpoints
• Follow RESTful API patterns consistent with other providers
• Implement standardized request/response formats
• Include comprehensive error handling and status code management
• Maintain compatibility with existing Deepchat workflows
• Implemented core features first (chat, images, TTS) followed by RAG/Agents
• Prioritized consistency with existing provider patterns over custom optimizations

Additional Information

Core Endpoints:

1. Chat Completions (POST /chat/completions)

• OpenAI Compatible
• Supports multiple AI models
• Streaming and non-streaming responses
• System prompts and conversation history

2. Models List (GET /models)

• OpenAI Compatible
• Available model discovery
• Model capability filtering
• Real-time model availability
• Detailed model information (context length, capabilities, etc.)

3. Image Generation (POST /images/generation)

• Text-to-image synthesis
• Multiple image sizes and quality options
• Batch generation support
• Ideogram and other advanced model support

4. Text-To-Speech (POST /tts/create)

• Input text to speech audio
• ElevenLabs multilingual model
• Voice id through GET tts/elevenlabslist

5. Video Generation (POST /image/tovideo)

• Image-to-video
• Custom parameters
• Multiple formats

6. RAG Query (POST /rag)

• Context-aware responses
• Knowledge bases
• Real-time retrieval

7. RAG Documents (GET /rag/user)

• Browse documents
• Filtering
• Pagination

8. Agent Create (POST /agent)

• Create agents
• Configurable roles
• Tool integration

9. Agent List (GET /agent)

• List agents
• Filtering
• Real-time status

Authentication:

• Bearer token authentication using API keys
• Secure credential storage following existing patterns
• Error handling for invalid or expired tokens

Error Handling:

• Comprehensive HTTP status code mapping
• Rate limit awareness and retry logic
• Network timeout management
• Graceful degradation when services are unavailable

Fruitful integration examples:

• https://github.com/Enconvo/straico
• https://alterhq.com/docs#external-api-keys
• https://straico.com/integrations