Kimi K3 Arrives on AWS Bedrock with 1M Token Context
AWS has made Kimi K3, a 2.8 trillion parameter open-weight model from Moonshot AI, available on Amazon Bedrock. The model features native vision capabilities, a 1-million-token context window, and support for explicit prompt caching. It is positioned for long-running coding and knowledge work tasks that require sustained context across large documents and repositories.
TL;DR
- Kimi K3 is now available on Amazon Bedrock with 2.8 trillion parameters and native vision capabilities
- The model offers a 1-million-token context window and approximately 2.5x improvement in scaling efficiency over Kimi K2
- Kimi K3 is the first open-weight model on Bedrock to support explicit prompt caching, reducing latency and input costs
- AWS has added dozens of open-weight models to Bedrock since 2025, with platform-level capabilities like tool calling and structured output
Why It Matters
Open-weight models are shifting AI economics by allowing organizations to match workloads to the right balance of capability, speed, and cost. Kimi K3's large context window and efficiency gains make it practical for sustained work on complex coding and knowledge tasks. The availability of such models on managed platforms like Bedrock lowers barriers to production deployment.
Business Impact
Companies can now adopt high-capability open models without building custom infrastructure or changing security practices. AWS's data boundary protections, zero data retention, and zero operator access mean teams can use Kimi K3 for sensitive work while maintaining compliance and data control. The model's prompt caching feature directly reduces inference costs for repeated context usage.
Key Implications
- Open-weight models are becoming viable alternatives to proprietary models for enterprise workloads, particularly in coding and knowledge work
- Managed platforms like Bedrock are consolidating access to multiple open models with consistent security and API standards, reducing vendor lock-in concerns
- The 1-million-token context window enables new use cases around long-document analysis and large codebase understanding that were previously impractical
What to Watch
Monitor whether Kimi K3 adoption accelerates on Bedrock and whether other open-weight model providers follow with similarly large context windows. Watch for performance comparisons between Kimi K3 and proprietary models on coding and knowledge tasks. Track whether prompt caching becomes a standard feature across other open models on Bedrock.
Subscribe to the newsletter
The latest stories and analysis, delivered to your inbox.
Free. No spam. Unsubscribe any time.
