Pull to refresh
Logo
What Changed
Following Why Sign Up
Moonshot AI's Kimi K3 goes live on Amazon Bedrock

Moonshot AI's Kimi K3 goes live on Amazon Bedrock

New Capabilities

The 2.8-trillion-parameter open-weight model adds India-only cloud processing as Moonshot releases a Pro successor.

6 days ago: Moonshot releases Kimi K3.1 Pro open weights

Overview

Updated 6 days ago

Amazon Bedrock now runs Moonshot AI's Kimi K3, a 2.8-trillion-parameter open-weight model, through global and India-only inference profiles. The India option, added November 3, keeps requests inside the Mumbai and Hyderabad regions for customers with data-residency rules. Kimi K3 has native vision, a one-million-token context window, and explicit prompt caching, the first open-weight model on Bedrock with that feature.

Moonshot released the open weights for a newer model, Kimi K3.1 Pro, on November 5. It has the same 2.8 trillion total parameters, about 104 billion active per request, and a one-million-token context window. Bedrock's model card still lists K3, not K3.1 Pro; AWS says customer data stays inside its boundary and is not used to train the model.

Why it matters

Chinese open-weight models can now run inside US and Indian cloud regions, giving enterprises a way to use them without exporting data.

Questions about this story

Free account needed to ask — your question is kept and asked for you right after sign-up. Answers are public.

No questions yet — be the first to ask.

Key Indicators

2.8T
Model parameters
The first open model to reach 2.8 trillion parameters.
1M
Context window
Tokens of context Kimi K3 can process in a single request, supporting long coding sessions and multi-document analysis.
2.5x
Scaling efficiency vs Kimi K2
Moonshot's claimed improvement from the new Kimi Delta Attention architecture.
16/896
Active experts per token
Experts activated per request under the Stable LatentMoE framework.
2
India destination regions
Mumbai and Hyderabad are the only regions used by the India geographic inference profile.

Voices

Curated perspectives — historical figures and your fellow readers.

Sign up to generate historical perspectives on this story.

Log in Sign Up

Organizations Involved

Moonshot AI
Moonshot AI
AI company
Released Kimi K3.1 Pro open weights on November 5; Kimi K3 remains the model on Amazon Bedrock.

Chinese developer of the Kimi family of language models, including Kimi K3 and the newer Kimi K3.1 Pro.
Amazon Web Services (AWS)
Amazon Web Services (AWS)
Cloud computing provider
Hosts Kimi K3 on Amazon Bedrock through global and India geographic inference profiles.

Amazon's cloud division, offering Kimi K3 through Bedrock with global and India-only inference profiles and explicit prompt caching.

Timeline

July 2026 November 2026

7 events Latest: 6 days ago
Tap a bar to jump to that date

  1. Moonshot releases Kimi K3.1 Pro open weights

    Latest Product Launch

    Moonshot publishes full weights for Kimi K3.1 Pro on Hugging Face. The multimodal model has 2.8 trillion total parameters, about 104 billion active per request, and a one-million-token context window.

  2. AWS adds India-only inference for Kimi K3

    Partnership

    Amazon Bedrock adds an India geographic inference profile for Kimi K3, keeping processing in the Mumbai and Hyderabad regions.

  3. Kimi K3 goes generally available on Amazon Bedrock

    Partnership

    Kimi K3 becomes generally available on Amazon Bedrock through cross-Region inference profiles from a range of AWS Regions, with explicit prompt caching as an open-model first on the platform.

  4. Full Kimi K3 weights released

    Product Launch

    Moonshot releases complete Kimi K3 model weights under the Kimi K3 License, letting other organizations host and modify the model.

  5. Demand overwhelms Moonshot's capacity

    Operational

    Moonshot pauses new subscriptions to its hosted Kimi K3 service three days after launch as demand exceeds available computing power.

  6. Moonshot AI unveils Kimi K3

    Product Launch

    Moonshot AI reveals Kimi K3, a 2.8-trillion-parameter open-weight model with native vision and a 1-million-token context window.

  7. Moonshot revises commercial licence for Kimi K3

    Business

    The updated licence requires firms hosting Kimi K3 as a service to enter a paid arrangement once aggregate annual revenue exceeds $20 million.

Scenarios

1

Moonshot lands US cloud revenue-sharing deals

Possible Resolves by Q2 2027

Discussed by: Bloomberg Intelligence analysts cited via Quartz; reporting by finance.biggo.com on the negotiations

Moonshot is seeking up to 30% of revenue from Kimi K3 services hosted on Microsoft Azure, Amazon Web Services, and Google Cloud, according to people familiar with the talks. AWS already hosts the model on Bedrock. Whether Microsoft or Google follow depends on unresolved questions about revenue split, data access, and how token usage is audited for billing.

2

Microsoft Azure or Google Cloud follows AWS with Kimi K3 hosting

Possible Resolves by Q1 2027

Discussed by: finance.biggo.com reporting on early-stage negotiations between Moonshot and Microsoft, Amazon, and Google

Moonshot has held preliminary talks with Microsoft and Google about hosting Kimi K3 on Azure Foundry and Vertex AI, but no agreements are guaranteed. If either cloud announces general availability, the model reaches every major US hyperscaler. If both decline, AWS becomes Kimi K3's only major US distribution channel.

3

US regulators impose restrictions on Chinese open-weight models

Unlikely Resolves by Q3 2027

Discussed by: insideai.news raised scrutiny questions; Noah Intelligence highlighted the geopolitical divide in cloud deployments

Hosting a Chinese-developed model inside US cloud infrastructure may draw regulatory attention. AWS has not disclosed any special review of Kimi K3 beyond standard model onboarding. New Commerce Department or CFIUS guidance targeting Chinese open-weight models in US clouds could reshape this arrangement and affect how other Chinese models reach US enterprises.

4

AWS adds Kimi K3.1 Pro to Amazon Bedrock

Possible Resolves by Feb 5, 2027

Discussed by: Moonshot's November 5 open-weights release and AWS's existing K3 hosting

Moonshot published K3.1 Pro weights on November 5, but Bedrock's model card still lists K3. If AWS adds the newer model, Bedrock customers get the successor's expanded coding and agent training without changing clouds. If AWS waits, K3.1 Pro users must self-host or use another provider.

Historical Context

3 moments from history that rhyme with this story — and how they unfolded.

January 2025

DeepSeek R1 (2025)

Chinese lab DeepSeek released an open-weight reasoning model that matched OpenAI's top-tier reasoning model on several benchmarks at a fraction of the cost. The release triggered a global selloff in US AI stocks, with Nvidia losing hundreds of billions in market value in one session.

Then

US lawmakers introduced bills to restrict Chinese AI; US tech companies scrambled to explain the competitive gap.

Now

Established that Chinese labs can produce frontier open-weight models, reshaping US-China AI competition.

Why this matters now

Kimi K3 is the next major Chinese open-weight frontier model, but unlike DeepSeek R1, it now has enterprise distribution inside a US hyperscaler rather than relying on direct downloads.

July 2023

Meta Llama 2 on AWS Bedrock (2023)

Meta released Llama 2 as an open-weight model and made it available through AWS Bedrock and Microsoft Azure. It was among the first major deployments of an open-weight model on enterprise cloud platforms with security controls attached.

Then

Open-weight models gained credibility for enterprise use, with major clouds providing hosting and governance.

Now

Set the pattern of open-weight models distributed through hyperscalers, later followed by Mistral, DeepSeek, and others.

Why this matters now

Kimi K3 extends this pattern at vastly larger scale and with a Chinese developer, testing how far open-weight distribution can go when geopolitical scrutiny is higher.

February 2024

Mistral on Microsoft Azure (2024)

French AI startup Mistral signed a deal with Microsoft to make its models available on Azure, with a $16 million investment from Microsoft. The deal gave Mistral global distribution while keeping its models open-weight.

Then

Mistral gained enterprise distribution through a US hyperscaler, though the investment drew European regulatory review.

Now

Showed that non-US open-weight model makers can reach enterprise markets through US clouds, but not without scrutiny.

Why this matters now

Kimi K3 is a similar arrangement with a Chinese developer, where the investment stakes, regulatory scrutiny, and geopolitical tension are all higher.

Sources

(18)
+12
Inside AI The AWS News Feed Dataforcee Digital Cloud Computing Köln Web Pulse GitHub (Moonshot AI official) BigGo Finance Noah Intelligence Scientific American CNN Business Amazon Web Services South China Morning Post ZeroWiz Magazine AWS Builder Center AI Affairs AI Top News Amazon Web Services Inside AI