According to CoinMeta, AWS announced that Grok 4.7 has now been launched on Amazon Bedrock, featuring a cutting-edge model designed specifically for coding and long-duration tasks. This model provides a context window of 500k and supports four configurable levels of inference effort: low, medium, high, and ultra-high. Grok 4.7, at the bedrock-runtime endpoint, enables cross-regional inference configuration, supporting responses, chat completion, and API. According to XAI, this is their most powerful model in the fields of coding and knowledge work; it can handle difficult tasks for longer periods and verifies its outputs more meticulously before proceeding.