ElysiumAI gateway поддерживает OpenAI, Anthropic, Gemini и rout.my-compatible aliases на одном PAYG-провайдере.
OpenAI: https://your-domain.com/v1
Anthropic: https://your-domain.com/anthropic/v1
Gemini: https://your-domain.com/compatible/v1beta
Chat Completions, Responses, embeddings, images, audio, rerank, moderations and realtime use the /v1 base URL.
curl https://your-domain.com/v1/chat/completions \
-H "Authorization: Bearer sk-your-ElysiumAI-key" \
-H "Content-Type: application/json" \
-d '{"model":"gpt-4.1-mini","messages":[{"role":"user","content":"Hello"}]}'Claude SDKs can use the Rout-style /anthropic/v1/messages alias, or the legacy /v1/messages route.
curl https://your-domain.com/anthropic/v1/messages \
-H "x-api-key: sk-your-ElysiumAI-key" \
-H "anthropic-version: 2023-06-01" \
-H "Content-Type: application/json" \
-d '{"model":"claude-3-5-sonnet-latest","max_tokens":256,"messages":[{"role":"user","content":"Hello"}]}'Gemini SDKs can point to /compatible/v1beta. The native /v1beta alias is kept too.
curl "https://your-domain.com/compatible/v1beta/models/gemini-2.5-flash:generateContent?key=sk-your-ElysiumAI-key" \
-H "Content-Type: application/json" \
-d '{"contents":[{"parts":[{"text":"Hello"}]}]}'Any OpenAI-compatible SDK can point to the ElysiumAI /v1 base URL.
import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.ElysiumAI_API_KEY,
baseURL: "https://your-domain.com/v1",
});
const response = await client.chat.completions.create({
model: "gpt-4.1-mini",
messages: [{ role: "user", content: "Hello" }],
});| Area | Method | Route | Purpose |
|---|---|---|---|
| OpenAI | GET | /v1/models | List models |
| OpenAI | GET | /v1/models/{model} | Retrieve model |
| OpenAI | POST | /v1/chat/completions | Chat Completions |
| OpenAI | POST | /v1/completions | Legacy Completions |
| OpenAI | POST | /v1/responses | Responses API |
| OpenAI | POST | /v1/responses/compact | Responses compaction |
| Anthropic | POST | /anthropic/v1/messages | Rout-style Messages alias |
| Anthropic | POST | /v1/messages | Legacy Messages alias |
| Anthropic | GET | /anthropic/v1/models | Anthropic model list |
| Gemini | GET | /compatible/v1beta/models | Rout-style Gemini model list |
| Gemini | POST | /compatible/v1beta/models/{model}:{action} | generateContent, streamGenerateContent, embedContent, batchEmbedContents |
| Gemini | GET | /v1beta/models | Native Gemini alias |
| Gemini | POST | /v1beta/models/{model}:{action} | Native Gemini relay |
| Embeddings | POST | /v1/embeddings | OpenAI embeddings |
| Embeddings | POST | /v1/engines/{model}/embeddings | Engine-style embeddings |
| Images | POST | /v1/images/generations | Image generation |
| Images | POST | /v1/images/edits | Image edits |
| Images | GET | /embed/image:{model} | Raw image bytes for HTML/Markdown |
| Audio | POST | /v1/audio/transcriptions | Speech to text |
| Audio | POST | /v1/audio/translations | Translate audio |
| Audio | POST | /v1/audio/speech | Text to speech |
| Video | POST | /v1/video/generations | Async video generation |
| Video | GET | /v1/video/generations/{task_id} | Fetch video task |
| Video | POST | /v1/videos | OpenAI-style video create |
| Video | GET | /v1/videos/{task_id} | OpenAI-style video retrieve |
| Video | GET | /v1/videos/{task_id}/content | Proxy video content |
| Video | POST | /kling/v1/videos/text2video | Kling text to video |
| Video | POST | /kling/v1/videos/image2video | Kling image to video |
| Other | POST | /v1/rerank | Rerank |
| Other | POST | /v1/moderations | Moderations |
| Other | GET | /v1/realtime | OpenAI Realtime WebSocket |
| Tasks | POST | /mj/submit/imagine | Midjourney-compatible tasks |
| Tasks | POST | /suno/submit/{action} | Suno-compatible tasks |
For HTML or Markdown image embedding, pass the API key in the query string because browsers cannot attach Authorization headers to image tags.
<img src="https://your-domain.com/embed/image:gpt-image-1?key=sk-your-ElysiumAI-key&prompt=marine%20gateway" alt="generated" />
| Area | Route | Purpose |
|---|---|---|
| Profile | /profile | Own profile, password and 2FA |
| Analytics | /dashboard | Balance, spend graph and recent requests |
| Top Up | /dashboard/top-up | Wallet balance and payment history |
| PAYG | /pricing | Wallet top-ups and model pricing |
| Admin | /admin | Providers, models, users, logs, payments and platform settings |
401/403 means the API key, session or role is not valid for the route.
429 means request, model or global rate limits are active.
PAYG usage is charged from wallet quota and audited in /api/log/self and admin ledger views.
Payment top-up settings are served by GET /api/user/topup/info.