Anthropic - input tokens + `max_tokens` > context length

[krissetto]

When using the Anthropic provider and reaching the upper limits of the available context length, we can end up in scenarios where the input tokens + the max_tokens value is > than the max supported context length for a given model.

To fix this, we need to keep in consideration the length of the input and set an appropriate max_tokens value before sending requests to the Anthropic API

Example error

400 Bad Request

{
  "type": "error",
  "error": {
    "type": "invalid_request_error",
    "message": "input length and `max_tokens` exceed context limit: 178959 + 64000 > 200000, decrease input length or `max_tokens` and try again"
  },
  "request_id":"req_011CUB2XUHWkFnQzxcKBbjpF"
}