Context, not generation
Brand truth in context. Fewer wasted tokens.
Encoded compiles and serves brand context for other tools to use. We are not a content generator. When that context is cached, model providers can spend fewer input tokens on repeated brand instructions.
Anthropic prompt caching can reduce input costs by up to 90% for eligible cached prompts. That figure is Anthropic's, with model, cache, and pricing conditions. It is not an Encoded customer outcome.
The BCP package is markdown files plus Registry delivery over HTTPS and MCP. Better content and fewer revision rounds come from giving agents one living source, not from Encoded writing the copy.
Attribute the up-to-90% caching figure to Anthropic. Do not present it as an Encoded guarantee.
Read live pricing