One key, every model
Text, image, video, speech and embeddings from the major labs, behind one endpoint and one bill.
One key to every model the world ships.
REDH Token
Every model,
one API
Routing that fails over
Each call goes to the cheapest or fastest healthy channel. When one stalls, the next one answers.
Drop-in for your stack
Keep the OpenAI, Claude or Gemini SDK you already use. Change the base URL and the key.
Every model, one line.
REDH Token connects the models, meters every call and settles one bill.
Format bridging
Write in OpenAI format and call Claude or Gemini. Each request is rewritten for the upstream, and the answer comes back in your format.
Prompt caching
Long system prompts, tool definitions and documents that repeat are read from cache and billed at a fraction of the input price.
Line-by-line ledger
Every call is weighed when it ends: prompt, cached and completion tokens, each at its own rate, then matched against the upstream bill.
One line to every lab
From Hong Kong to the Bay Area, Paris, Beijing and Hangzhou
Labs are grouped by the region of their headquarters.