Alibaba Launches Wan3.0 AI Video Model Days After $10.2 Billion Share Sale to Fund AI Infrastructure

Reviewed byNidhi Govil

5 Sources

Share

Alibaba unveiled Wan3.0, its latest AI video generation model capable of creating 30-second clips from diverse inputs including PDFs and PowerPoint files. The launch follows a record $10.2 billion Hong Kong share placement, with all proceeds earmarked for AI capabilities. Priced competitively against Google Veo, the model targets corporate and marketing use cases while raising questions about profitability amid soaring AI infrastructure costs.

Alibaba Wan3.0 Launches With Enhanced Video Generation Capabilities

Alibaba officially launched Wan3.0, its latest AI video generation model, on August 24, 2026, just one day after completing a $10.2 billion share placement in Hong Kong—the largest primary follow-on offering by a Hong Kong-listed company.

1

2

The timing signals Alibaba's commitment to translating massive AI infrastructure spending into tangible commercial products that investors can evaluate. Developed by Alibaba's Tongyi Lab, the 30-second video model represents a significant leap from its predecessor Wan2.7, which maxed out at 15 seconds per clip.

3

Source: The Next Web

Source: The Next Web

Wan3.0 had been running in public beta since August 6, 2026, through Alibaba Cloud's Model Studio and Qwen Cloud platform, making Monday's announcement an expansion of access rather than a debut.

1

The AI video model generates clips at resolutions up to 1080p while maintaining character detail, spatial consistency, props, and motion graphics throughout the full length.

1

It renders faces with synchronized micro-expressions and delivers multilingual voice output, positioning it as a business tool for corporate communications teams and marketing departments rather than primarily serving filmmakers.

1

Competitive API Pricing Targets Enterprise Users

Alibaba set API pricing for Wan3.0 at $0.05 per second for 480p output, $0.10 for 720p, and $0.20 for 1080p—translating to $12 per minute at the highest resolution.

2

3

This pricing strategy significantly undercuts Google Veo 3.1's standard tier, which costs $0.40 per second.

1

However, several rivals ranking above Alibaba's previous model on independent leaderboards also cost less per minute than Wan3.0 at 1080p resolution.

1

An Indian startup recently shipped an AI-powered video generation model at half a cent per second, betting that most commercial video production doesn't require frontier quality.

1

The competitive pricing reflects broader market dynamics where Chinese labs have quietly built commanding positions in AI video generation while Western attention remained focused on chatbots.

1

Yet Wan3.0 hasn't been independently benchmarked, making quality claims company-issued for now and leaving buyers to weigh price against unverified performance.

1

Document-to-Video Conversion Targets Corporate Use Cases

What distinguishes Wan3.0 from competitors is its input flexibility. Beyond text, images, audio, and video, the AI video model accepts web pages and documents including PDFs and PowerPoint files.

1

4

Users can hand the model a slide deck or spreadsheet and receive a video reflecting the source material's structure.

1

This capability points directly at marketing departments and corporate communications teams, though Alibaba lists short-film production, advertising, tourism promotion, and music video creation among use cases.

4

5

The model also supports simulation training for autonomous vehicles and robotics.

1

Access remains controlled rather than open. The public beta runs through Model Studio and Qwen Cloud on an application basis, full API availability is described as rolling out, and a consumer-facing site at wan.video is promised as a members-only platform.

1

This staged rollout suggests Alibaba is metering demand while calculating compute costs for serving the model at scale.

Closed-Weight Models Mark Strategic Shift From Open Source

Notably, Wan3.0's weights are closed, marking a departure from Alibaba's reputation for open-sourcing the Wan series.

1

Wan2.2, released in July 2025, remains the last video flagship the company released as open source.

1

This shift toward closed-weight models aligns with Alibaba's decision to start charging the heaviest users of its open Qwen models, signaling a broader monetization strategy as AI infrastructure costs mount.

1

Financial Pressure Mounts Despite Revenue Growth

The commercial rationale showed up clearly in Alibaba's first-quarter fiscal 2027 results disclosed August 20, 2026. Cloud and AI revenues grew 45% year-over-year to 48.44 billion yuan, the fastest pace in 22 quarters.

2

3

AI-related product revenues reached 12.4 billion yuan, marking 12 consecutive quarters of triple-digit growth and lifting the annualized run rate above 49.5 billion yuan.

2

AI model services now carry more than 16 billion yuan in annualized recurring revenue.

1

However, quarterly net profit plunged 75% as capital expenditure surged 75% to 67.68 billion yuan.

1

3

The AI Cloud and Computing Services segment posted adjusted EBITA of 5.63 billion yuan, up 133%, with margin expanding to 12%.

2

Free cash flow turned negative at 44.7 billion yuan.

2

Alibaba committed 380 billion yuan to AI infrastructure over three years in early 2025 and has been reported weighing an increase to 480 billion yuan.

1

3

Investor Concerns Surface Despite AI Investment Narrative

Alibaba shares dropped 8.1% to HK$113 in Hong Kong trading following the Alibaba share sale announcement, reflecting investor concerns over the dilutive impact of issuing 710 million new shares at an 8.4% discount.

2

The stock movement underscored tension between near-term shareholder dilution and longer-term AI infrastructure buildout that proceeds are meant to support.

2

All net proceeds from the share sale are earmarked for Alibaba's full-stack AI capabilities spanning chips, AI infrastructure, and model development.

2

Management has pointed to accelerating cloud growth and margin expansion as evidence that spending translates into commercial traction, positioning Wan3.0 as an early monetizable output of broader AI investment.

2

Whether the 30-second video model justifies this spending remains unanswered. Video models are expensive to run, buyers switch based on price, and independent leaderboards that would settle quality arguments haven't evaluated Wan3.0 yet.

1

The company's success will depend on attracting and retaining users in a rapidly evolving landscape where the AI race intensifies and competitors emerge from unexpected markets.

3

Watch for independent benchmarks, enterprise adoption rates, and whether Alibaba can sustain margin expansion while scaling compute-intensive video generation services.

Today's Top Stories

© 2026 TheOutpost.AI All rights reserved