AI Models
AI Video Generation Statistics 2026: ArtFlame Model and Workflow Index
A dated first-party snapshot of AI video model routes, audio-capable models, text-to-video, image-to-video, video-to-video, and guided workflows available in ArtFlame.
As of August 23, 2026, the ArtFlame production catalog enables 34 video model IDs and exposes 122 public guided Arcana workflow routes. This first-party snapshot documents what the catalog contains, how the counts were produced, and what they do not prove about output quality.
34
Enabled video model IDs
Thirty-seven supported IDs minus three routes marked coming soon.
23
Enabled audio-capable IDs
Catalog routes marked for audio after excluding coming-soon entries.
122
Public Arcana workflow routes
Canonical guided-workflow paths in the public SEO route manifest.
33
Text-to-video entries
Distinct entries in the Video Studio text-to-video catalog.
32
Image-to-video entries
Distinct entries in the Video Studio image-to-video catalog.
9
Video-to-video entries
Distinct entries in the Video Studio video-to-video catalog.
How this snapshot was counted
The counts come from ArtFlame production source manifests, not a survey or an estimate. Enabled video IDs are the supported video set after removing entries explicitly marked coming soon. Text-to-video, image-to-video, and video-to-video counts come from their separate Video Studio catalogs. Guided workflow routes come from the canonical Arcana SEO route manifest.
A model can appear in more than one creation mode, so the text-to-video, image-to-video, and video-to-video totals should not be added together as a count of unique models. The snapshot date matters because providers, versions, access, and controls change.
What the catalog breadth means
A multi-model catalog lets creators keep the brief, references, aspect ratio, and production context together while changing the generation route. That is useful when one shot needs reference fidelity, another needs audio, and a third needs a lower-cost draft.
More models do not automatically create better work. The practical advantage is the ability to match the model to the shot and compare outputs without rebuilding the whole project.
- Use text to video when the concept can begin from language alone
- Use image to video when composition or identity needs a visual anchor
- Use video to video when motion or timing already exists
- Use an Arcana workflow when the desired outcome is clearer than the model choice
What these numbers do not prove
Catalog counts do not prove that every model is equally reliable, fast, affordable, or appropriate for every account. They also do not measure customer satisfaction, generation success rate, or subjective visual quality.
ArtFlame is building a separate aggregate performance dataset for timing and reliability. Those figures should only be published after the production database connection is healthy, the sample is large enough, and the aggregation excludes private user content.
A better way to compare AI video models
Use the same reference, action, camera instruction, duration, and output target across candidate models. Record prompt adherence, subject stability, camera behavior, usable seconds, audio fit, generation time, credits, and repair work.
This produces a decision for a real production brief rather than a universal leaderboard. Update the result whenever a provider changes the model version or controls.
Citation and reuse
You may cite these figures as the ArtFlame catalog snapshot dated August 23, 2026. Link to this page so readers can see the methodology and update date.
Do not restate the numbers as market-wide adoption, market share, or performance statistics. They describe the ArtFlame production catalog only.
Frequently asked questions
How many AI video models does ArtFlame support?
The August 23, 2026 production snapshot contains 37 supported video model IDs, of which 34 are enabled and three are marked coming soon.
How many ArtFlame workflows are public?
The canonical Arcana SEO manifest contains 122 public guided-workflow routes in this snapshot.
Does the largest model catalog produce the best video?
No. Catalog size measures choice, not quality. Compare candidate models against the same brief, references, duration, and evaluation criteria.
Are these market-wide AI video statistics?
No. These are first-party ArtFlame catalog statistics. They do not measure market share, adoption, or every AI video product.