1. Start
  • Start
    • Product Intrduction
    • Quick to use
    • Using Nexhina in AI Coding Tools
  • API integration explanation
    • Quick Start
    • Getting an API Key
    • Authentication
    • Request URL
    • HTTP Status Codes
    • Streaming Output Guide
  • API Endpoints
    • Chat
      • Chat Completion
    • Models
      • List Available Models
    • Responses
      • Responses API
    • Embeddings
      • Text Embedding
    • Images
      • Text-to-Image
      • Image-to-Image
    • Audio
      • Text-to-Speech (TTS)
      • Speech-to-Text (STT)
      • Audio Translation
    • Video
      • Generate Video
    • Moderation
      • Content Moderation
    • Rerank
      • Rerank
  • Platform-related
    • Platform agreement
    • Privacy Policy
    • General Questions
  1. Start

Product Intrduction

About Nexhina#

Nexhina is a unified AI gateway that delivers OpenAI-compatible access to the world's leading models — through a single endpoint, one API key, and one bill.

What problems does it solve?#

Pain PointNexhina's Solution
Juggling multiple providers (OpenAI, Anthropic, Google, Mistral, DeepSeek…)One OpenAI-compatible endpoint for every model — no SDK switching
Fragmented billing, quotas, and contracts per vendorA single dashboard, a single invoice, and a unified credit pool across all models
Inconsistent latency and reliability across regionsGlobally distributed edge with intelligent routing — closest node wins
Vendor lock-in when prices change or models get deprecatedSwitch models with a single string change, no code refactor
Compliance headaches (GDPR, EU AI Act, data residency)Regional data routing, EU representative, and audit-ready logs
Re-implementing auth, retries, streaming, and tool-calling per providerAll standard OpenAI features work out of the box — Function Calling, Vision, JSON Mode, streaming

Core Capabilities#

Standard Interface, Zero Lock-In: Fully OpenAI / Anthropic-compatible. Change base_url, keep your existing SDK and code unchanged.
All Major Models in One Place: GPT-4o, Claude Sonnet 4, Gemini 2.5 Pro, o3, plus a curated selection of regional models — covering Chat, Embedding, Image, Audio, Video, Moderation, and Rerank (11 interfaces in total).
Global Edge Routing: Requests are automatically routed to the lowest-latency region. Data residency options for EU, US, and APAC.
Unified Multi-Key Management: Create as many API keys as you need. Each key has its own quota, model allowlist, rate limit, and usage analytics.
Token-Level Billing: Pay only for what you consume. No monthly minimums, no seat fees.
Enterprise-Ready: SSO/SAML, audit logs, role-based access, and private deployment options.

One-Line Integration#

Supported Tools#

Cursor, Windsurf, Continue, JetBrains AI, VS Code Copilot, ChatBox, LobeChat, NextChat, Open WebUI, Dify, LangChain, LlamaIndex — and any other tool that speaks the OpenAI API standard.

Who is it for#

Individual developers who want a single key to experiment with every frontier model
Startups that need to move fast without signing five separate vendor contracts
Enterprises that require centralized governance, audit trails, and predictable billing across business units
Multi-region teams that need consistent performance from Singapore to Frankfurt to São Paulo

Modified at 2026-07-03 01:41:04
Next
Quick to use
Built with