Anthropic has suspended access to Fable 5 to comply with a US government export-control directive, so we've disabled claude-fable-5 on RedRoute. Requests to it now time out and are not billed — you won't be charged for any failed call. All other models are unaffected. We'll re-enable routing as soon as access is restored.
AI gateway
The AI gateway that cuts your API bill
RedRoute is an OpenAI-compatible AI gateway, router, and proxy. One endpoint routes to every major model — GPT-5.5, Claude Opus 4.8, DeepSeek V4 Flash, DeepSeek V4 Pro, MiniMax M3, GLM-5.1, Qwen3.7 Max — with automatic fallback and a lower effective rate. Same request, same response shape, lower line item.
# Drop-in: change one line. No SDK rewrite, no code changes.
from openai import OpenAI
client = OpenAI(
api_key="rr_live_...",
base_url="https://redroute.pro/v1" # <- the only change
)
resp = client.chat.completions.create(
model="gpt-5.5", # or claude-opus-4-8, deepseek-v4-pro, ...
messages=[{"role":"user","content":"hi"}]
)
What you get
OpenAI-compatible
No rewrite
Same /v1 endpoints and response shapes as the OpenAI API. Point your existing SDK at one base URL and you're done.
One router
Every model, one key
Route GPT, Claude, DeepSeek, MiniMax, GLM and Qwen through a single unified API. Switch models by changing a string.
Resilient
Automatic fallback
When a provider rate-limits or goes down, requests reroute automatically so your app stays up — no on-call scramble.
Cheaper
Lower effective rate
Pay below the official per-token price on the models you already use, billed the same way. The savings show up on day one.
What you actually pay
FAQ
What is an AI gateway?
An AI gateway is a single API in front of many model providers. Instead of integrating OpenAI, Anthropic and others separately, you call one endpoint that handles routing, fallback, and billing — so swapping or combining models is a config change, not a rewrite.
Is RedRoute OpenAI-compatible?
Yes. The endpoints and response shapes match the OpenAI API, so any OpenAI SDK or library works by changing only the base URL and key.
Which models can I route?
GPT-5.5, Claude Opus 4.8, DeepSeek V4 Flash, DeepSeek V4 Pro, MiniMax M3, GLM-5.1, Qwen3.7 Max, plus image and audio models — all through the same gateway.
How much can I save versus calling the providers directly?
RedRoute charges an effective rate below the official per-token price on supported models, with the exact discount shown per model on the comparison page.
Related: model price comparison · full model catalog · api docs · pricing & plans