Skip to content

Sociaro AI Gateway

The Sociaro gateway is a single entry point to models from many vendors — OpenAI, Anthropic, Alibaba, ByteDance, DeepSeek, Zhipu AI, xAI, Google and more — with one API key, unified usage analytics, cost attribution, budgets and rate limits managed for you.

Across these providers the catalog spans text, vision/omni, image and video models — flagships like Qwen, Seed, DeepSeek, Grok and GLM (text), Seedream, Qwen-Image and Z-Image (image), Wan, Seedance and HappyHorse (video). Fetch the full, current list with GET /v1/models (or /gw/models for metadata).

Base URL: https://api.sociaro.com

You call the provider’s native API, just with a different host and your gateway key. The gateway is byte-transparent: request and response bodies are forwarded unmodified (including SSE streaming), so every provider SDK, cookbook example and API reference works as-is.

https://api.openai.com/v1/chat/completions
↓ becomes ↓
https://api.sociaro.com/openai/v1/chat/completions

The gateway:

  1. authenticates your gw_live_… key and checks it is allowed to use the requested model;
  2. swaps in the real provider credential server-side — you never handle provider keys;
  3. forwards the request and streams the response back byte-for-byte;
  4. meters usage and cost in the background (never on your request path).

Get a key from your organization admin (or the dashboard), then:

Terminal window
curl https://api.sociaro.com/openai/v1/chat/completions \
-H "Authorization: Bearer gw_live_..." \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-4o",
"messages": [{"role": "user", "content": "Hello!"}]
}'

Anthropic uses its native header, exactly like the upstream API:

Terminal window
curl https://api.sociaro.com/anthropic/v1/messages \
-H "x-api-key: gw_live_..." \
-H "anthropic-version: 2023-06-01" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-sonnet-4-6",
"max_tokens": 256,
"messages": [{"role": "user", "content": "Hello!"}]
}'

Or use the Python SDK:

from sociaro_ai import Sociaro
client = Sociaro(api_key="gw_live_...")
response = client.chat.completions.create(
model="openai/gpt-4o",
messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

Or the TypeScript SDK:

import { Sociaro } from "sociaro-ai";
const client = new Sociaro({ apiKey: "gw_live_..." });
const response = await client.chat.completions.create({
model: "openai/gpt-4o",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(response.choices[0].message.content);

Because the proxy is byte-transparent, official SDKs work by overriding the base URL and key:

# OpenAI SDK
from openai import OpenAI
client = OpenAI(api_key="gw_live_...", base_url="https://api.sociaro.com/openai/v1")
# Anthropic SDK
from anthropic import Anthropic
client = Anthropic(api_key="gw_live_...", base_url="https://api.sociaro.com/anthropic")
  • One key, many providersauthentication & scopes
  • Usage & cost analytics — per provider, model, day, or your own attribution tags; see GET /gw/stats
  • Exact billing via the SDK — transparent proxy meters approximately and reconciles monthly; the SDK meters every call (incl. streams) exactly — see billing & usage accuracy
  • Async media jobs (video) with accurate cost finalization — async jobs
  • Self-service key management — issue and revoke keys for your team without contacting support — api-reference
  • Reseller tooling — sub-accounts with per-customer budgets and markup, plus a white-label end-customer portal — api-reference
  • Stripe Token Billing — export per-client usage to Stripe Billing meters and invoice your customers per token — billing-stripe
  • Budgets and rate limits — predictable spend; clear 402 / 429 error contract
  • Evidence export & retention — tamper-evident audit trail for auditors and procurement — evidence
  • EU compliance — Frankfurt residency, eu_only region routing, 185-day retention — eu-compliance

A web dashboard with usage charts, key and sub-account management, budgets, CSV export and async-job status is available for your organization. Ask your admin for access.