Moonshot AI releases the Kimi K3.1 Pro Highspeed model on the Amazon Bedrock platform
Moonshot AI has released the Kimi K3.1 Pro Highspeed model on Amazon Bedrock – a faster version of Kimi K3 with a 1M token window, focused on agentic tasks, with data isolation within AWS and a choice of pricing tiers.
Moonshot AI has released the Kimi K3.1 Pro Highspeed model on the Amazon Bedrock platform. It is a post-trained version of the Kimi K3 model with the same 3T-class architecture, open weights, and a 1 million token context window, trained with substantially more post-training compute. According to Moonshot AI, this additional training is focused on long-running agentic work and reducing factual errors, and it also improves tasks such as financial modeling, office documents, and programming for game engines, 3D scenes, and CAD scripts.
The model generates output tokens faster; according to Moonshot AI, it reaches 2–3 times the speed compared to the Kimi K3.1 Pro model. Kimi K3.1 Pro Highspeed uses the same endpoint, API, and features as Kimi K3 on Amazon Bedrock, so users of existing code can try the model by simply changing the model ID. It supports tool calling, structured output, a configurable reasoning level (reasoning_effort), image input, and streaming.
Three profiles are available for the model for routing requests between regions: a global profile (cheaper per token, suitable for most workloads) and regional profiles that keep processing within the US or India for cases with data residency requirements. According to AWS, data remains within the AWS data boundary, is not shared with Moonshot AI, and is not used for further training of the model; zero data retention is always active for inference, and AWS operators do not have access to prompts and outputs.
Amazon Bedrock also offers a Flex tier for the model at half the price of the standard tier (at the cost of potentially longer response times, suitable e.g. for model evaluation or batch tasks), explicit prompt caching for repeated context that reduces latency and input costs, and integration with the open source coding agent OpenCode via an OpenAI-compatible API.
Why it matters
For developers, this is a practically risk-free upgrade of existing integrations with Kimi K3 on Amazon Bedrock – higher speed and, according to the manufacturer, fewer factual errors without needing to change code. For companies, the important combination is lower latency for agentic and coding tasks together with documented data protection guarantees (zero retention, no operator access, choice of regional data residency in the US or India) and the ability to optimize costs via the Flex tier or prompt caching.
Release card
Kimi K3.1 Pro Highspeed
Moonshot AI
- Specifications
- 3T-class architecture, the same as in the Kimi K3 model.
- Context
- 1 million tokens
- Inputs
- Text and image input, text output; supports tool calling, structured output, and streaming.
- Availability
- Available on the Amazon Bedrock platform via the Chat Completions, Responses, Converse, and Invoke API as well as via Playground, with profiles for the global, US, and Indian regions, and via integration with the OpenCode tool.
- Rychlost generování výstupních tokenů 2 až 3krát vyšší než u modelu Kimi K3.1 Pro According to Moonshot AI, this model generates responses faster than the previous version, Kimi K3.1 Pro.
- Long-running agentic tasks
- Financial modeling and work with office documents
- Programming for game engines, 3D scenes, and CAD scripts
A faster follow-up version of Moonshot AI's Kimi K3.1 Pro model, focused on long-running agentic tasks and reducing factual errors.
The card summarizes information from the article and any dated corrections, with a link to the original source. It is not our assessment of the model. It does not yet have a dedicated editorial profile. Model selection and other announcements →
Two audiences, two different impacts
What this means
For individuals
Developers who already call Kimi K3 via Amazon Bedrock can switch to K3.1 Pro Highspeed without changing the API, just by adjusting the model ID, and can influence speed/price using the reasoning_effort parameter or by choosing the Flex tier.
For a business
Companies building agentic systems or coding assistants on Amazon Bedrock get a faster and, according to Moonshot AI, more accurate model with preserved data isolation (zero data retention, no access by AWS operators, no data sharing with Moonshot AI) and a choice of regional data residency (US/India), which is relevant for compliance requirements.
DevelopmentCheck the original
Event sources
clearly official source · 1 publisher, 0 independent. We count feeds from the same owner only once.