Moonshot AI’s Kimi K3.1 Pro Highspeed is now available through Amazon Bedrock. Developers already using Kimi K3 can try the new version by changing the model identifier while keeping the same endpoint and API structure.
The model supports tool calling, structured output, image input, streaming and adjustable reasoning effort. AWS offers an OpenAI-compatible Chat Completions and Responses interface alongside its Converse and Invoke APIs. A global inference profile can route work among supported commercial regions, while US and India profiles keep processing within those locations for data-residency needs.
AWS says prompts and completions remain inside its data boundary, are not shared with Moonshot AI or used for model training, and use zero data retention. The Flex tier costs half the Standard inference price but may respond more slowly, while explicit prompt caching can reduce repeated-context costs. Moonshot claims the Highspeed edition generates output two to three times faster than Kimi K3.1 Pro; developers should validate that vendor-reported gain on their own prompts and latency requirements.