Generative AI Market Statistics: Size, Growth & Forecasts
Generative AI market 2026: ~$67B total (up 36% YoY), projected $1.3T by 2032 (Bloomberg Intelligence). Image generation segment ~$11.2B, video & audio ~$6.8B. Enterprise genAI spending reached $37B in 2025.
Generative AI is the fastest-growing technology sector ever measured: from roughly $8 billion in 2022 to an estimated $67 billion in 2026, with Bloomberg Intelligence projecting $1.3 trillion by 2032. Visual generation - images plus video and audio - accounts for roughly $18 billion of the 2026 total and is growing faster than the text segment.
Key statistics
The global generative AI market is estimated at $67 billion in 2026, a ~36% increase over 2025’s $44.8B.
Bloomberg Intelligence and Goldman Sachs project the generative AI market reaching $1.3 trillion by 2032 at a 36% CAGR.
Image generation is the second-largest genAI segment at ~$11.2B in 2026, behind text/NLP ($31.4B) and ahead of code generation ($8.6B).
Enterprise generative AI spending alone reached $37 billion in 2025.
The video & audio generation segment is estimated at $6.8B in 2026, projected to reach $165B by 2032.
Generative AI market by segment, 2026 versus 2032
Text still dominates with $31.4B of the $67B total, but visual media compounds faster: image generation is projected to grow ~18x to $210B by 2032, versus ~18x for text off a much larger base.
| Segment | 2026 size | 2032 projection | Share of 2026 market |
|---|---|---|---|
| Text & NLP (LLMs) | $31.4B | $580B | ~47% |
| Image generation | $11.2B | $210B | ~17% |
| Code generation | $8.6B | n/a | ~13% |
| Video & audio | $6.8B | $165B | ~10% |
| Other (3D, agents, etc.) | ~$9B | n/a | ~13% |
Segment breakdown from SearchLab’s 2026 analysis; total market and 2032 projections from Grand View Research and Bloomberg Intelligence.
How does the generative AI market break down by segment?
Text and NLP still lead, but visual media (image + video/audio) is the fastest-compounding slice - projected to grow from ~$18B in 2026 to ~$375B by 2032.
LLMs remain the largest segment, projected to reach $580B by 2032.
Image generation is projected to grow from $11.2B (2026) to $210B by 2032 - an ~18x increase.
IDC reports 80% of Fortune 500 companies use OpenAI’s generative AI in some capacity.
The generative AI market was valued at ~$59B in 2025 growing at a 37.57% CAGR, targeting $400B by 2031 in mid-range models.
How we compiled this data
We anchored the headline number to the most-corroborated 2026 estimate (~$67B, consistent across Grand View and SearchLab) and used Bloomberg Intelligence for the long-range projection because it is the most-cited institutional model. Segment splits come from a single source (SearchLab) and are labeled accordingly; enterprise spending comes from separate IDC-derived reporting. Last full review: June 12, 2026.
Before you cite these numbers
- The $1.3T-by-2032 projection dates to Bloomberg Intelligence’s 2023 model. It remains the standard citation but has not been re-validated against 2025–2026 actuals.
- Segment boundaries are blurry: multimodal models (GPT-4o, Gemini) generate text, images, and code from one product, making clean segment accounting partly arbitrary.
- The "80% of Fortune 500 use OpenAI" stat measures any usage, including single-team pilots; it is not an enterprise-wide deployment figure.
Frequently asked questions
How big is the generative AI market in 2026?
Approximately $67 billion, up ~36% from 2025. Projections put the market at $400 billion by 2031 (conservative) to $1.3 trillion by 2032 (Bloomberg Intelligence).
What share of generative AI is image and video?
Image generation accounts for roughly $11.2 billion and video & audio for $6.8 billion in 2026 - together about 27% of the total market, and growing faster than text.
How much are enterprises spending on generative AI?
Enterprise generative AI spending reached $37 billion in 2025, with 80% of Fortune 500 companies using OpenAI technology alone.
Sources
Figures on this page are compiled from the following publishers and reports. Where sources disagree, we present the range and note the methodology difference.