One API for simpler model access and configuration.
Keep your existing official SDKs and swap only the base_url to access connected Claude, GPT, Gemini, and other models with centralized routing, usage, logs, and billing.
Choose a model from the live catalog to see a matching code example.The Viral Pelican Benchmark
Why is the global AI community using 'a pelican riding a bicycle' to stress-test LLMs? Avian anatomy combined with mechanical bicycle kinematics immediately exposes spatial hallucinations, joint misalignments, and reasoning degradation. Copy the prompt below to test any model:
Layered routing architecture
Scroll to explore the routing, resilience, and billing layers. The cards illustrate platform concepts, not live system state.
Unified model access worldwide
A unified API for users worldwide. Request speed depends on your network, the Superoute gateway, and the status of model services such as GPT, Claude, and Gemini.
Upstream errors and routing
When an upstream is rate limited or unavailable, the gateway returns the corresponding error. Clients can configure retries based on the error type.
Transparent token billing and audit logs
Usage is organized by calendar day and organization. The console supports custom quotas, request-level diagnostics, and streaming token accounting; consult the console for recorded charges.
3-Step Drop-in Setup · Zero Rewrite
No third-party proprietary client libraries. No prompt rewrite. Seamless side-by-side verification.
from openai import OpenAI
import os
client = OpenAI(
base_url="https://api.superoute.uk/v1", # replace the gateway URL api_key=os.environ["SUPEROUTE_API_KEY"], # shared API key)
response = client.chat.completions.create(
model="model-from-catalog", # use a model ID from the live catalog
messages=[{"role": "user", "content": "Hello World"}],
)Swap base_url host only
Keep your existing official OpenAI or Anthropic SDKs — just point the base URL to Superoute production edge.
Single Key, Frontier Model Access
Use one API key to access models currently available in the unified catalog.
Routing and Error Handling for Upstream Variability
The gateway handles routing and upstream responses; actual availability depends on the upstream provider and network conditions.
Use the native endpoint. Keep the authentic developer experience.
Use compatible endpoints and verify the features required by your client against the selected model.
Current GPT models listed in the model catalog
Current Claude models listed in the model catalog
Current Gemini models listed in the model catalog
Standard HTTP requests with zero client SDK dependency
Unified model routing
All API keys use the same model catalog and routing. Check the catalog for current availability and pricing.
Native Desktop App · Local Control & Agent Orchestration
Superoute Desktop works out of the box. Manage API keys, agent configs, usage analytics, and balance locally without editing configuration files by hand.
Transactional Rust configuration for Claude Code, Codex, OpenCode, and OpenClaw with automatic backup, validation, and rollback.
Create, rename, and revoke. Plaintext secret is shown once and never persisted in forms.
Authoritative model catalog, 7-day request logs, cost breakdowns, and instant recharge.
Gateway health check, today's spend & token summary, and rapid agent environment toggle.
Everything you want to know before connecting
Have more questions? Feel free to reach out to our engineering support.