Genspark2API provides a unified API gateway for various AI models, including large language models, image generation, and video generation. It aims to simplify access to these models for developers and users by offering a single interface, supporting features like streaming, web search, and multi-modal input.

How It Works

The project acts as a proxy, forwarding requests to different AI model providers. It handles authentication, request formatting, and response parsing. Key features include support for multiple LLM providers (GPT-4o, Claude, Gemini), image generation models (DALL-E 3, Ideogram), and video generation models. It also incorporates advanced functionalities like web search integration, multi-turn conversations with image/file context, and automatic cookie management for handling rate limits and service availability.

Quick Start & Requirements

• Install/Run: Deploy via Docker Compose (docker-compose pull && docker-compose up -d) or Docker (docker run ... deanxv/genspark2api:latest).
• Prerequisites: Requires Docker. Environment variables GS_COOKIE (user session cookie) and API_SECRET (optional API key for authentication) must be set. A proxy service (genspark-playwright-proxy) is recommended for Cloudflare V3 bypass.
• Setup: Deployment via Docker is straightforward. Setting up the proxy service and configuring cookies might take additional time.
• Docs: Genspark AI

Highlighted Details

• Supports streaming and non-streaming chat completions.
• Integrates web search by appending -search to model names.
• Handles image and file input for multi-turn conversations.
• Offers text-to-image and text/image-to-video generation capabilities.

Maintenance & Community

• Community discussion via Telegram group.
• Project is actively maintained by deanxv.

Licensing & Compatibility

• License not explicitly stated in the README.
• Designed for integration with platforms like NextChat and One-API.

Limitations & Caveats

• Official services may enforce ReCaptchaV3, requiring a separate proxy service (genspark-playwright-proxy) for validation.
• Reliance on user cookies means potential issues with cookie expiration or invalidation.
• "Mixture-of-Agents" mode is triggered for unsupported models.