Moonshot AI Closes $3.5B+ Series F at a $35B Valuation as Kimi’s Pre-IPO Round Starts Early

Introduction

Moonshot AI, the Chinese AI company behind Kimi, has reportedly completed a Series F financing round of more than $3.5 billion**, pushing its post-money valuation to approximately **$35 billion.

The pace of the fundraising is almost as notable as the size.

According to multiple Chinese financial-media reports on July 29, 2026, investor demand exceeded the company’s original fundraising target by more than three times. The Series F was therefore closed earlier than planned.

Moonshot AI had originally been expected to begin its next financing discussions in August. Instead, the company has reportedly already started its Series G, described by the reports as its Pre-IPO round, at a target pre-money valuation of $50 billion.

The company has not publicly issued a detailed financing announcement confirming the round size, investor list, or final Pre-IPO terms. These figures should therefore be treated as well-sourced media reports rather than company-filed transaction data.

The financing arrives immediately after the release and open-weight publication of Kimi K3, Moonshot AI’s 2.8-trillion-parameter multimodal model.

Series F Raises More Than $3.5 Billion

The central financing figure reported across AIBase, The Paper, 21st Century Business Herald, and other Chinese outlets is straightforward:

The round is unusually large for a private Chinese AI company.

Earlier reports had described the financing as being conducted at approximately a **$31.5 billion pre-money valuation**. The newly reported post-money valuation of roughly $35 billion is broadly consistent with that earlier framing once the new capital is included.

Because the final amount is described as “more than $3.5 billion,” the published valuation numbers should be understood as rounded transaction figures rather than exact cap-table calculations.

Oversubscription Forced an Early Close

The more interesting detail is the level of investor demand.

People familiar with the financing told Chinese financial media that subscriptions exceeded the target by more than three times.

Rather than continue accepting capital until the previously planned deadline, Moonshot AI reportedly closed the round early.

Oversubscription does not automatically prove that a company will justify its valuation in the long run. It does, however, show that available allocation in the round was significantly smaller than investor demand.

For Moonshot AI, the timing also matters.

The fundraising took place while Kimi K3 was attracting attention for:

• Its 2.8-trillion total parameter count.
• A one-million-token context window.
• Native multimodal capabilities.
• Long-horizon coding and agentic workflows.
• The release of full model weights.
• Rapid adoption across the open-model ecosystem.

The financing therefore came at a moment when both technical visibility and investor interest were unusually high.

The Pre-IPO Round Has Already Started

The next round had previously been expected to begin in August.

According to the latest reports, that schedule has moved forward.

Moonshot AI has reportedly already started discussions for its Series G / Pre-IPO round, with a target pre-money valuation of $50 billion.

People close to the fundraising told The Paper that demand is again strong and that the round could close relatively soon.

That does not mean the $50 billion financing has been completed.

At the time of writing:

• The Series F is reported as completed.
• The Series G / Pre-IPO process is reported as having started.
• The $50 billion figure is a target pre-money valuation.
• Final investors, proceeds, allocation, and closing terms have not been publicly confirmed by Moonshot AI.

This distinction is important because early fundraising discussions can change before legal closing.

Why the $35B and $50B Valuations Are Different

The two headline valuations refer to different stages.

$35 Billion: Series F Post-Money Valuation

The reported $35 billion figure describes Moonshot AI after the Series F capital is included.

$50 Billion: Series G / Pre-IPO Pre-Money Target

The reported $50 billion figure describes the value investors would assign to the company before new money from the next round is added.

The Pre-IPO valuation is therefore not simply another description of the completed Series F.

It represents a new financing step at a substantially higher proposed valuation.

Earlier Reports Had Already Pointed to a $50B Pre-IPO Round

The $50 billion figure did not first appear on July 29.

Bloomberg-linked reporting and Chinese financial outlets had already reported in the previous week that Moonshot AI planned to begin a final pre-IPO fundraising round in August with a target valuation of up to $50 billion.

Those earlier reports said:

• The company was completing a financing round at roughly a $31.5 billion pre-money valuation.
• A final outside round could follow shortly afterward.
• Moonshot AI was considering a Hong Kong listing.
• The Pre-IPO process could begin in August.

The July 29 reports update that timeline in two ways:

1. The Series F is now described as completed at a roughly $35 billion post-money valuation.
2. The planned Pre-IPO round is said to have started earlier than originally expected.

The new information therefore looks more like an acceleration of an existing financing plan than a completely new strategy.

Kimi K3 Arrived Just Before the Financing Accelerated

Moonshot AI introduced Kimi K3 in mid-July and released its full model weights on July 27.

The model is important to the financing story because it has become the company’s clearest technical flagship.

Moonshot AI describes Kimi K3 as:

• A 2.8-trillion-parameter model.
• An open-weight, native multimodal model.
• A model with a 1-million-token context window.
• A long-horizon coding and agentic model.
• A model built on Kimi Delta Attention (KDA) and Attention Residuals (AttnRes).
• A Mixture-of-Experts system using a Stable LatentMoE design.
• A model capable of working with text, images, and video.

The company says K3 activates 16 of 896 experts and uses architectural and training improvements intended to increase scaling efficiency.

These are Moonshot AI’s official model specifications and technical claims, separate from the financing reports.

Kimi K3 Is Fully Available as an Open-Weight Model

The complete Kimi K3 model repository is now public through Moonshot AI’s official GitHub organization and Hugging Face account.

The release includes:

• Model weights.
• Configuration files.
• Model code.
• A technical report.
• A model card.
• Deployment guidance.
• The Kimi K3 license.

Moonshot AI’s license allows broad use, modification, deployment, fine-tuning, and redistribution subject to its stated terms.

This open-weight strategy is relevant to the company’s valuation because it gives Kimi a path to adoption beyond Moonshot AI’s own hosted product.

Developers can:

1. Use the official Kimi API.
2. Run Kimi K3 through supported inference systems.
3. Integrate the weights into private or enterprise infrastructure.
4. Build agents and products around the model.
5. Adapt the model within the license terms.

That creates a larger ecosystem than a closed consumer chatbot alone.

The Model Is Designed for More Than Chat

Moonshot AI positions Kimi K3 around agentic work rather than only conversational question answering.

Its official launch material highlights several categories.

Long-Horizon Coding

K3 is designed to navigate large repositories, use terminal tools, and work across extended software-engineering sessions.

Moonshot AI has demonstrated the model on tasks including compiler development, GPU-kernel optimization, game development, CAD, and chip-design experiments.

Knowledge Work

The company also positions K3 for research and office-style workflows that combine:

• Search.
• Document analysis.
• Data processing.
• Visualization.
• Dashboards.
• Reports.
• Presentations.

Native Multimodality

K3 processes text, images, and video within the same model family.

This allows visual information to remain part of an agent’s workflow rather than being handled only through a separate image model.

The capabilities help explain why investor attention is not centered only on Kimi’s consumer chat product.

Moonshot AI is trying to compete at the foundation-model, developer-platform, agent, and enterprise levels at the same time.

Open Weights Can Support Commercial Growth Rather Than Prevent It

Open-weight releases are sometimes framed as being in conflict with commercial value.

In practice, the relationship is more complicated.

An open-weight model can create several routes to revenue:

• Official API usage.
• Enterprise deployments.
• Managed inference.
• Team products.
• Developer tools.
• Premium agent applications.
• Support and integration.
• Ecosystem adoption that drives users toward hosted services.

Moonshot AI’s official Kimi K3 page provides both downloadable weights and paid API access.

At the time reviewed for this article, the official API listed Kimi K3 pricing at:

The company says its inference architecture achieves high cache-hit rates in coding workloads.

API prices and product terms can change, so production users should verify the current pricing page before making cost assumptions.

Reported Revenue Growth Has Also Strengthened the Funding Story

Separate media reports citing people familiar with Moonshot AI have said the company’s annual recurring revenue reached approximately $300 million in June**, up from roughly **$100 million in March.

These figures have been reported by Bloomberg-linked outlets and other financial media, but they have not been independently audited in public financial statements.

They matter because AI valuations are increasingly being judged on more than model benchmarks.

Investors want evidence that model capability can translate into:

• API demand.
• Enterprise customers.
• Paid subscriptions.
• Developer adoption.
• Recurring revenue.

Moonshot AI’s reported revenue trajectory, combined with Kimi K3’s technical launch, helps explain why its financing discussions accelerated so quickly.

Consumer Demand Has Also Tested Kimi’s Capacity

After the K3 launch, Chinese media reported that demand pushed Moonshot AI’s infrastructure close to capacity.

The company temporarily prioritized existing paid users rather than continuing to expand new consumer subscriptions at the same pace.

This is a useful reminder that model popularity creates infrastructure costs.

A frontier-model company needs capital not only for research, but also for:

• Training runs.
• Inference clusters.
• Long-context serving.
• Storage.
• Networking.
• Model deployment.
• Enterprise reliability.
• Engineering teams.

A multi-billion-dollar financing round can therefore support both the next model and the infrastructure required to serve the current one.

From a $20B Round to $35B in a Few Months

Moonshot AI’s valuation has changed rapidly during 2026.

In May, TechCrunch and Forbes reported that the company was raising approximately $2 billion** at a valuation around **$20 billion.

By June and July, reports placed the company’s next financing at a much higher valuation.

The latest Series F report now puts Moonshot AI at around $35 billion post-money**, while the Pre-IPO round is targeting a **$50 billion pre-money valuation.

The speed of that progression is unusual even by frontier-AI standards.

It also creates a higher bar.

As valuations rise, investors will expect the company to continue demonstrating:

• Model quality.
• Product adoption.
• Revenue growth.
• Infrastructure execution.
• Competitive differentiation.
• A credible route toward a public-market listing.

What “Pre-IPO” Means Here

The term Pre-IPO does not guarantee that an initial public offering will occur on a fixed schedule.

It usually refers to a private financing round conducted when a company is already actively considering or preparing for a public listing.

Reports have said Moonshot AI may pursue a Hong Kong IPO and has shared listing-related proposals with investors.

The company itself had not publicly confirmed a final IPO timetable in the sources reviewed for this article.

A Pre-IPO round can still be followed by:

• Additional private financing.
• A delayed listing.
• A changed venue.
• Revised valuation.
• A restructuring process.

Readers should therefore distinguish “Pre-IPO fundraising” from “IPO approved.”

What Is Confirmed and What Is Reported

The financing story combines several types of evidence.

Keeping those categories separate prevents a fast-moving fundraising story from turning preliminary deal terms into facts.

Why Investors Are Paying Attention to Moonshot AI

The original AIBase report is primarily about financing rather than a broader market analysis.

Still, the funding can be understood through three immediate factors.

1. A Frontier Model Launch

Kimi K3 gave investors a fresh technical event to evaluate.

1. Open-Weight Distribution

The full model can spread through developers, cloud providers, enterprises, and independent inference platforms.

1. Reported Commercial Growth

Media reports of rapidly rising recurring revenue give investors a business metric alongside the technical story.

Together, those factors create a stronger fundraising narrative than model performance alone.

Risks Behind a Rapidly Rising Private Valuation

A large financing round does not remove the basic risks of the frontier-model business.

Moonshot AI still needs to manage:

• Very high compute costs.
• Fast-moving competition.
• Model commoditization.
• Price pressure in inference.
• Infrastructure shortages.
• Enterprise reliability expectations.
• The cost of operating extremely large open-weight models.
• Public-market scrutiny if an IPO proceeds.

Kimi K3’s 2.8-trillion-parameter scale is also demanding to deploy.

Moonshot AI’s own documentation recommends large supernode configurations with 64 or more accelerators for efficient serving.

An open model may be freely downloadable while still being expensive to run at full scale.

What to Watch Next

Final Pre-IPO Terms

The most immediate question is whether the reported $50 billion target is maintained when the Series G closes.

Investor List

The identity and allocation of investors will show which strategic and financial institutions are backing Moonshot AI at this stage.

IPO Preparation

Any formal filing, restructuring announcement, exchange documentation, or underwriter appointment would provide stronger evidence of the listing timetable.

Kimi K3 Adoption

GitHub, Hugging Face, API usage, enterprise deployment, cloud integration, and developer activity will show whether K3’s launch translates into durable adoption.

Revenue Quality

If reported ARR continues growing, investors will want to know how much comes from repeatable enterprise and API usage rather than short-term launch demand.

Kimi K4

Moonshot AI is expected to keep pushing its model roadmap. The next generation will need to justify the much higher valuation expectations now being attached to the company.

常见问题

How much did Moonshot AI raise in its latest funding round?

Multiple Chinese financial-media reports say Moonshot AI completed a Series F of more than $3.5 billion. The company’s reported post-money valuation is approximately $35 billion.

Is Moonshot AI really valued at $50 billion?

Not from the completed Series F. The $50 billion figure is the reported pre-money target for the company’s newly started Series G / Pre-IPO round.

Why did Moonshot AI close the Series F early?

People familiar with the financing told Chinese media that investor subscriptions exceeded the original target by more than three times. The strong demand reportedly led Moonshot AI to close the round ahead of schedule.

Has Moonshot AI completed its Pre-IPO round?

No completed Series G transaction has been publicly confirmed in the sources reviewed here. Reports say discussions began earlier than planned and could move quickly, but final terms have not been announced.

Is Moonshot AI planning an IPO?

Media reports say the company is preparing for a possible Hong Kong listing and describes the Series G as a Pre-IPO round. Moonshot AI has not publicly announced a final listing date.

What is Kimi K3?

Kimi K3 is Moonshot AI’s open-weight 2.8-trillion-parameter multimodal model. It supports a one-million-token context window and is designed for long-horizon coding, agentic knowledge work, reasoning, and visual tasks.

Are Kimi K3 weights available to download?

Yes. Moonshot AI released the full weights through its official GitHub and Hugging Face repositories under the Kimi K3 License.

Why is Kimi K3 relevant to Moonshot AI’s valuation?

The model gives Moonshot AI a current frontier product, an open-weight developer ecosystem, and a paid API offering. Investors are also evaluating reported revenue growth and whether K3 adoption can translate into durable commercial demand.

相关工具

• Kimi: Moonshot AI’s main agentic AI workspace for chat, research, documents, coding, and multimodal tasks.
• Kimi API Platform: Official API access for Kimi K3 and other Moonshot AI models.
• Kimi Code: Moonshot AI’s terminal and IDE coding agent for software-engineering workflows.
• Kimi K3 on GitHub: Official model repository containing the technical report, model information, code, and license.
• Kimi K3 on Hugging Face: Official open-weight model page for downloading and deploying Kimi K3.
• vLLM: An open-source inference engine supported in Moonshot AI’s Kimi K3 deployment guidance.

Related Links

• The Paper: Moonshot AI Series F and Early Pre-IPO Round: Reporting on the $3.5B+ Series F, $35B post-money valuation, oversubscription, and accelerated Series G.
• 21st Century Business Herald: Moonshot AI Financing: Additional reporting on the completed Series F and early start of the Pre-IPO round.
• Kimi K3 Official Technical Blog: Moonshot AI’s official introduction to Kimi K3, its architecture, capabilities, availability, and pricing.
• Kimi K3 Official GitHub Repository: Official open-weight release, technical report, and model documentation.
• Kimi K3 Hugging Face Model Page: Official weights, model card, configuration, and deployment resources.
• TechCrunch on Moonshot AI’s May 2026 Funding: Earlier reporting on the company’s $2B financing at a $20B valuation.

Summary

Moonshot AI has reportedly completed a Series F of more than $3.5 billion at a post-money valuation of roughly $35 billion. The round was heavily oversubscribed and closed ahead of schedule.

The company’s next Series G / Pre-IPO round, previously expected to begin in August, is now reported to be underway with a target pre-money valuation of $50 billion. That round has not yet been publicly confirmed as completed.

The financing acceleration comes immediately after the launch and full open-weight release of Kimi K3, Moonshot AI’s 2.8-trillion-parameter multimodal model with a one-million-token context window.

The key development is not only the size of the Series F, but the speed at which Moonshot AI has moved from a $35 billion post-money round into negotiations around a $50 billion Pre-IPO valuation.