Marketing adoption of AI video generation jumped from 18% to 41% in a single year — and the technology behind that shift traces directly to Meta’s Make-A-Video, a research model released in September 2022. Make-A-Video was the first publicly demonstrated system to generate video from text without requiring paired text-video training data. This article covers the key Make-A-Video user data and statistics for 2026, including training infrastructure, market size, geographic adoption, and how the model fits into a rapidly changing competitive landscape.
Make-A-Video Statistics
- Meta trained Make-A-Video on 2.3 billion text-image pairs and 20 million unlabeled videos, launched in September 2022.
- The text-to-video generation market reached $614.8 million in 2024, with projections pointing to $2,562.9 million by 2032 at a 20.0% CAGR.
- Marketing adoption of AI video tools rose 128% year-over-year, from 18% in 2023 to 41% in 2024.
- Monthly active users across all AI video platforms topped 124 million in January 2026, with generation volume up 840% since January 2024.
- Meta committed $60–65 billion in AI capital expenditure for 2025, with plans to deploy over 1.3 million GPUs.
What Is Make-A-Video?
Meta unveiled Make-A-Video in September 2022 as a text-to-video generation system that uses two separate learning approaches: labeled image-text data to understand what the world looks like, and unlabeled video footage to understand how it moves. This separation removed the need for paired text-video training data, which had previously been a significant bottleneck for the field.
All videos generated by the system include a watermark to identify them as AI-produced content. As of early 2026, Meta has not released Make-A-Video for public use, keeping it in internal research and testing phases while developing successor models.
Make-A-Video Training Data
Make-A-Video uses a diffusion-based architecture with pseudo-3D convolutions and temporal attention mechanisms. The system generates 16-frame sequences at a native resolution of 64×64 pixels, upscaled to 768×768 pixels through super-resolution networks. Maximum output length is 5 seconds, with no audio generation capability.
| Dataset | Type | Volume |
|---|---|---|
| LAION (filtered) | Text-image pairs | 2.3 billion |
| WebVid-10M | Unlabeled video | 10 million |
| HD-VILA-100M (subset) | Unlabeled video | 10 million |
Source: Meta AI Research, Make-A-Video Paper (2022)
Meta applied content filtering to remove NSFW material, toxic text, and images with a watermark probability exceeding 0.5 before training.
Make-A-Video Market Size Statistics
The text-to-video generation category that Make-A-Video helped define was valued at $614.8 million in 2024. Research firms project the market will reach $2,562.9 million by 2032, at a 20.0% compound annual growth rate. The broader AI video generation platform market, which includes all tools from text-to-video to avatar-based systems, stood at $6.2 billion in 2025 and is expected to reach $7.77 billion in 2026.
Source: MarketIntelo, AI Video Generation Platform Market Report (Q4 2025); Fortune Business Insights AI Video Generator Market (2025)
| Year | Market Size (USD) | YoY Change |
|---|---|---|
| 2022 | $2.48 billion | — |
| 2023 | $3.51 billion | +41.5% |
| 2024 | $4.89 billion | +39.3% |
| 2025 | $6.20 billion | +26.8% |
| 2026 | $7.77 billion | +25.3% |
| 2030 | $19.17 billion | — |
| 2034 | $47.80 billion | — |
Source: MarketIntelo, AI Video Generation Platform Market Research Report (March 2026)
Make-A-Video User Adoption Statistics
Marketing adoption of AI video generation rose from 18% in 2023 to 41% in 2024, a 128% year-over-year increase. That figure continued climbing: 75% of video marketers reported using AI tools in their workflows as of 2025, and 78% of marketing teams now incorporate AI-generated video into at least one campaign per quarter.
Monthly active users across all AI video platforms reached 124 million in January 2026. AI video generation volume grew 840% between January 2024 and January 2026, according to data compiled by Vivideo. Among larger companies, 73% of Fortune 500 firms have integrated AI video tools into their content workflows.
Source: Wyzowl Video Marketing Report (2024); Vivideo AI Video Statistics 2026
| Metric | Figure |
|---|---|
| Marketing adoption (2023) | 18% |
| Marketing adoption (2024) | 41% |
| Video marketers using AI tools (2025–2026) | 75% |
| Fortune 500 companies using AI video | 73% |
| Marketing teams using AI video per quarter | 78% |
| AI-generated video content on social platforms | 40% |
| Monthly active users on AI video platforms (Jan 2026) | 124 million |
| AI video generation volume growth (Jan 2024 – Jan 2026) | 840% |
Source: Zebracat AI Video Creation Statistics (2025); Vivideo AI Video Statistics (2026)
Make-A-Video Production Efficiency Data
AI-generated video cuts average production costs by roughly 91%, from $4,500 per minute under traditional production to approximately $400 per minute. The time needed to produce a 60-second marketing video dropped from 13 days to 27 minutes with AI tools, according to Vivideo’s compiled industry benchmarks.
Organizations using AI video tools report efficiency gains of up to 80% in time and budget savings. The average marketing team saves 34 hours per week previously spent on production and editing. Companies using AI video also report 68% faster time-to-publish for video campaigns. For a broader look at how AI tools are reshaping creative workflows, the Claude statistics and usage trends article shows similar efficiency gains across text-based AI platforms.
Make-A-Video Statistics by Geography
North America leads the text-to-video market with a 40.61% share and $249.7 million in revenue in 2024. The United States alone is projected to account for $155.3 million in 2025. Asia-Pacific captured a 31.40% share in 2024 and is expanding at a comparable pace to North America, driven by extensive AI software adoption across marketing, education, and entertainment verticals.
Source: Fortune Business Insights, AI Video Generator Market Report (2025)
| Region | 2024 Market Share | 2024 Revenue | Projected CAGR |
|---|---|---|---|
| North America | 40.61% | $249.7 million | 20.3% |
| Asia-Pacific | 31.40% | — | 23.8% |
| Europe | ~20% | ~$123 million | ~19% |
| Rest of World | ~8% | — | — |
Source: Fortune Business Insights, AI Video Generator Market Report (2025)
Within Europe, Germany shows the highest growth rate at a projected 20.5% CAGR, partly driven by the German government’s commitment to allocate 5 billion euros toward AI development by 2025. France is projected to record $30.2 million in 2025, and the United Kingdom $26.8 million.
How Make-A-Video Compares to Other AI Video Models
Make-A-Video launched into a sparse competitive field in 2022. Within the same year, Google published Imagen Video (11.6 billion parameters) and Tsinghua University released CogVideo (9.4 billion parameters). These models showed different architectural approaches but all produced short video clips from text prompts.
By 2024, Meta’s own Movie Gen model surpassed Make-A-Video’s architecture substantially — scaling to 30 billion parameters for video generation and a separate 13 billion for audio, and extending maximum video length from 5 seconds to 16 seconds at 16 frames per second. Platforms like HeyGen AI, which serves over 40,000 paying businesses, demonstrate how commercial AI video tools have moved far beyond research-stage limitations.
Source: Meta AI Research; arXiv (2022–2024)
| Model | Developer | Parameters | Max Duration |
|---|---|---|---|
| Make-A-Video | Meta | Not disclosed | 5 seconds |
| CogVideo | Tsinghua University | 9.4 billion | N/A |
| Imagen Video | 11.6 billion | N/A | |
| Movie Gen (video) | Meta | 30 billion | 16 seconds |
| Movie Gen (audio) | Meta | 13 billion | — |
Source: Meta AI Research; arXiv preprints (2022–2024)
Meta Investment and Make-A-Video Infrastructure Statistics
Meta invested $39.2 billion in capital expenditure for AI development in 2024. For 2025, the company projected that figure would rise to $60–65 billion, alongside deployment of more than 1.3 million GPUs to support AI workloads. These numbers reflect how much of Meta’s broader AI infrastructure investment flows through the video generation research pipeline that Make-A-Video started.
Meta AI reached 1 billion monthly active users in Q1 2025. The introduction of the Vibes feed on September 25, 2025 — which brought short-form AI-generated video to the Meta AI mobile app — correlated with a sharp increase in daily app downloads, compared to just 4,000 daily downloads recorded a year earlier in October 2024.
Source: Meta Platforms Q4 2024 Earnings Report; Meta 2025 Guidance
FAQ
Was Make-A-Video ever released to the public?
No. Meta announced Make-A-Video in September 2022 but kept it in internal research and testing through early 2026. It has not been made publicly available, with Meta continuing to develop successor models like Movie Gen instead.
How was Make-A-Video trained?
The model trained on 2.3 billion filtered text-image pairs from the LAION dataset and 20 million unlabeled videos from WebVid-10M and HD-VILA-100M — without requiring any paired text-video data, which was its defining technical contribution.
What model replaced Make-A-Video at Meta?
Meta’s Movie Gen model, introduced in 2024, scaled to 30 billion parameters for video and 13 billion for audio, extending maximum output from Make-A-Video’s 5 seconds to 16 seconds at 16 frames per second.
How large is the text-to-video market in 2026?
The broader AI video generation platform market reached $6.2 billion in 2025 and is projected to hit $7.77 billion in 2026, expanding at a 25.4% compound annual growth rate through 2034, according to MarketIntelo.
How many people use AI video platforms in 2026?
Monthly active users across all AI video platforms surpassed 124 million in January 2026. AI video generation volume grew 840% between January 2024 and January 2026, per Vivideo’s industry data.
As a senior analyst, I benchmark and review gadgets and PC components, including desktop processors, GPUs, monitors, and storage solutions on Aboutchromebooks.com. Outside of work, I enjoy skating and putting my culinary training to use by cooking for friends.