CoolShow AniVerseShort animation drama intelligence, data & resources

INDUSTRY RESOURCES

Video models

Browse by role, industry stage, territory and freshness. Cards start with what it is, why it matters and what changed.

All75Companies12People3Models16Tools5Workflows7Services3Platforms6Titles11Markets3Policy3Events3Research3

16 public resources

Basic records and deep dossiers are labelled separately · compare up to four of one type
ModelsSources checked

LTX-2.3

What it is
LTX's previous API generation, retaining text, image, audio, retake, extend and reframe endpoints; new projects should compare it field by field with the LTX 2.5 matrix.
Why it matters
LTX's previous API generation, retaining text, image, audio, retake, extend and reframe endpoints; new projects should compare it field by field with the LTX 2.5 matrix. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 9 fields pending
ModelsSources checked

LTX 2.5

What it is
An audiovisual generation release with native multishot, automatic duration, portrait/landscape and output up to 4K; current 2.5 endpoints should not inherit 2.3 retake, extend or reframe assumptions.
Why it matters
An audiovisual generation release with native multishot, automatic duration, portrait/landscape and output up to 4K; current 2.5 endpoints should not inherit 2.3 retake, extend or reframe assumptions. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 9 fields pending
ModelsSources checked

Gemini Omni 1.1 Flash

What it is
A Google video-generation model update documented in August 2026, with up to ten seconds of prior context, extension in ten-second increments to forty seconds, boundary-frame interpolation, 360p previews and 4K upscaling. These are public capabilities, not short-drama acceptance-rate or cost findings.
Why it matters
Omni 1.1 Flash belongs in a short-drama shot test because it connects low-resolution preview, scene extension, boundary frames and 4K upscaling. The dossier does not celebrate maximum duration; it identifies the character, action, sound, cost and mastering gates that remain.
Recent change
Google publishes the Omni 1.1 Flash update
Production · Distribution · ResearchDeep dossier
ModelsSources checked

MiniMax Hailuo 2.3

What it is
A text- and image-to-video family whose official API documents standard and Fast variants, 768P/1080P, six/ten-second options and camera-motion commands.
Why it matters
Hailuo 2.3 combines standard, Fast and API routes. Short-drama teams need separate records for model ID, motion and micro-expression, camera commands, duration and resolution, then use accepted shots and repair work to explain any speed difference.
Recent change
Rechecked the Hailuo 2.3 release material and image-to-video API documentation
Production · Distribution · ResearchDeep dossier
ModelsSources checked

Luma Ray3.14

What it is
A text-, image- and video-to-video model whose official materials document native 1080p, HDR, EXR, boundary frames and multiple aspect ratios.
Why it matters
Ray3.14 should be treated as a general shot-generation route. Official materials document resolution, HDR/EXR, boundary frames and aspect ratios, but not native audio or project-level proof of complex motion, character continuity or cost per accepted shot.
Recent change
Rechecked the Ray3.14 release notes and video-model field guide
Production · Distribution · ResearchDeep dossier
ModelsSources checked

Luma Ray3.2 Modify

What it is
A production video-transformation model requiring source video, preserving duration, supporting up to 64 indexed keyframes and handling clips up to 20 seconds.
Why it matters
A production video-transformation model requiring source video, preserving duration, supporting up to 64 indexed keyframes and handling clips up to 20 seconds. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 9 fields pending
ModelsSources checked

Seedance 2.5

What it is
An audio-video joint generation model for up-to-30-second storytelling, with stronger reference understanding, audiovisual editing, white-model control and green-screen editing.
Why it matters
Seedance 2.5 matters to short-drama teams because it combines audiovisual generation, up-to-30-second storytelling, reference-video understanding and finer editing controls. Adoption still depends on fixed character, dialogue, action and delivery samples, not a vendor leaderboard.
Recent change
Official page rechecked: Seedance 2.5 still documents 30-second storytelling, reference and editing capabilities
Production · Distribution · ResearchDeep dossier
ModelsSources checked

Wan2.2 T2V / I2V

What it is
Wan2.2's foundational text- and image-to-video routes, including A14B mixture-of-experts models and a lighter 5B route across 480p and 720p generation.
Why it matters
Wan2.2's foundational text- and image-to-video routes, including A14B mixture-of-experts models and a lighter 5B route across 480p and 720p generation. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 9 fields pending
ModelsSources checked

Adobe Firefly Video

What it is
Adobe's video model for generating short clips in the Firefly timeline from text, boundary frames, composition or motion references; some controls are mutually exclusive and need shot-level configuration locks.
Why it matters
Adobe's video model for generating short clips in the Firefly timeline from text, boundary frames, composition or motion references; some controls are mutually exclusive and need shot-level configuration locks. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 10 fields pending
ModelsSources checked

Kling AI Video 3

What it is
A video-generation entry documented by Kling AI's official guide. The profile records public controls without using vendor self-evaluation as a cross-model ranking.
Why it matters
A video-generation entry documented by Kling AI's official guide. The profile records public controls without using vendor self-evaluation as a cross-model ranking. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 10 fields pending
ModelsSources checked

Runway Aleph

What it is
An in-context video editing model for adding or removing objects, changing environments and lighting, generating viewpoints and transferring style in existing footage.
Why it matters
An in-context video editing model for adding or removing objects, changing environments and lighting, generating viewpoints and transferring style in existing footage. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 10 fields pending
ModelsSources checked

HunyuanVideo 1.5

What it is
A video generation model with open weights, inference code, and text-to-video and image-to-video integration paths.
Why it matters
A video generation model with open weights, inference code, and text-to-video and image-to-video integration paths. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 10 fields pending
ModelsSources checked

Wan2.2-S2V

What it is
An open character-animation model that drives speaking, singing and performance video from an image and audio.
Why it matters
An open character-animation model that drives speaking, singing and performance video from an image and audio. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 10 fields pending
ModelsSources checked

Runway Gen-4.5

What it is
A video generation model whose official release documents both capability direction and known causal-reasoning and object-permanence limitations.
Why it matters
A video generation model whose official release documents both capability direction and known causal-reasoning and object-permanence limitations. This is currently a basic index record; no deep assessment is published until type-specific fields are complete.
Recent change
Awaiting the next source review
Production · Distribution · ResearchBasic record · 10 fields pending