Giotto.ai / giotto/giotto-alto

Giotto Alto - access through LLMTR

Giotto Alto is Giotto.ai's reasoning model. It can use its own tools, such as reading web pages and running code, before it answers, so a reply takes from a few seconds to a minute or more and arrives in one piece. It takes text and images (base64 or https) and returns text. reasoning_effort accepts medium (the default) and high; high checks the answer before returning it and uses about twice the tokens. Every request also carries about 8,000 input tokens of the model's own instructions, and each step the model takes is billed, so read the cost from the usage of each response. Function tools, response_format and stop are not supported; temperature, top_p and max_tokens are not applied, so the answer length cannot be capped. Memory, calendar, mail and document tools tied to a personal account are not available, and a request that tries to use them is refused.