2026 releases · AI video

    Every AI video model released in 2026

    96 models from 13 providers, dated 6 January 2026 to 25 August 2026. They did not arrive evenly — they landed in 16 bursts, and this page is those bursts in order.

    The 2026 video timeline

    6 January 2026

    5 models · 1 provider

    LTX shipped 5 video models on 6 January 2026. All 5 have a spec page. Cheapest complete job in the window: 20 credits on LTX 2. Highest output in it: 4K on LTX 2.

    • New name on the roster: LTX.
    ModelBuilt byCredits per jobMax outputClip lengths
    LTX 2 Retake VideoRegenerate and improve video segments with LTX 2LTX24-80 credits—3s, 5s, 10s
    LTX 2Latest video generation with enhanced controlLTX20-256 credits4K6s, 8s, 10s, 12s, 14s, 16s
    LTX 2 ProProfessional image-to-video with LTXV2LTX29-192 credits4K6s, 8s, 10s
    LTX 2 Text to Video FastFast text-to-video generation with LTXV2LTX20-256 credits4K6s, 8s, 10s, 12s, 14s, 16s
    LTX 2 Text to Video ProProfessional text-to-video with LTXV2LTX29-192 credits4K6s, 8s, 10s

    26–30 January 2026

    6 models · 2 providers

    Pixverse and Vidu shipped 6 video models between 26 January 2026 and 30 January 2026. Of the 6, 4 have a spec page and 2 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 6 credits on Vidu Q3 Turbo Image to Video. Highest output in it: 1080p on Pixverse 5.6 Image to Video.

    ModelBuilt byCredits per jobMax outputClip lengths
    Pixverse 5.6 Image to VideoLatest Pixverse model for converting images to high-quality animated videosPixverse28-240 credits1080p5s, 8s, 10s
    Pixverse 5.6 Text to VideoGenerate videos directly from text prompts with Pixverse 5.6 advanced capabilitiesPixverse28-240 credits1080p5s, 8s, 10s
    Vidu Q3 Image to VideoTransform static images into dynamic videos with Vidu Q3 technologyVidu14-93 credits1080p5s, 10s, 15s
    Vidu Q3 Turbo Image to VideovariantThe fast tier of Vidu Q3 at half the price: animates a still image with native audio, dialogue and sound effects. Any length from…Vidu6-99 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Vidu Q3 Turbo Text to VideovariantThe fast tier of Vidu Q3 at half the price: text to video with native audio, dialogue and sound effects. Any length from 1 to 16…Vidu6-99 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Vidu Q3 VideoGenerate high-quality videos from text prompts with Vidu Q3 engineVidu28-185 credits1080p5s, 10s, 15s

    1–5 February 2026

    19 models · 2 providers

    ByteDance and Kling shipped 19 video models between 1 February 2026 and 5 February 2026. Of the 19, 9 have a spec page and 10 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 20 credits on DreamActor V2. Highest output in it: 4K on Kling Video V3 4K Image to Video.

    ModelBuilt byCredits per jobMax outputClip lengths
    DreamActor V2AI-powered face reenactment model that transfers facial expressions and head movements from a driving video to a reference imageByteDance20-480 credits——
    Kling O3 Pro Image to VideoKling O3 Pro image-to-video with advanced reasoning-enhanced generation and camera controlsKling56-168 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling O3 Pro Reference to VideovariantKling O3 Pro reference-to-video for character-consistent generation using reference imagesKling56-168 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling O3 Pro Text to VideovariantKling O3 Pro text-to-video with reasoning-enhanced generation, camera controls, and audioKling56-168 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling O3 Pro Video to Video EditvariantKling O3 Pro video-to-video editing with reasoning-enhanced prompt-based modificationsKling68 credits——
    Kling O3 Pro Video to Video ReferencevariantKling O3 Pro video-to-video reference-based editing for style and character transferKling68-202 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling O3 Standard Image to VideoKling O3 Standard image-to-video with reasoning-enhanced generationKling45-135 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling O3 Standard Reference to VideoKling O3 Standard reference-to-video for character-consistent generationKling45-135 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling O3 Standard Text to VideovariantKling O3 Standard text-to-video with reasoning-enhanced generationKling45-135 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling O3 Standard Video to Video EditvariantKling O3 Standard video-to-video editing with prompt-based modificationsKling51-1210 credits——
    Kling O3 Standard Video to Video ReferencevariantKling O3 Standard video-to-video reference-based editing for style transferKling51-152 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling Video V3 4K Image to VideovariantKling V3 4K image-to-video - cinema-grade resolution with advanced camera controls, audio generation, and element compositionKling168-504 credits4K3s, 4s, 5s, 6s, 7s, 8s
    Kling Video V3 4K Text to VideoKling V3 4K text-to-video - cinema-grade resolution with advanced camera controls, audio generation, and multi-prompt supportKling168-504 credits4K3s, 4s, 5s, 6s, 7s, 8s
    Kling Video V3 Pro Image to VideovariantKling V3 Pro image-to-video generation with camera controls, audio generation, and element compositionKling68-202 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling Video V3 Pro Motion ControlKling V3 Pro motion control - transfer motion from a driving video to a reference image at pro-tier qualityKling68-1613 credits1080p—
    Kling Video V3 Pro Text to VideoKling V3 Pro text-to-video generation with advanced camera controls, audio generation, and multi-prompt supportKling68-202 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling Video V3 Standard Image to VideoKling V3 Standard image-to-video generation with camera controls and element compositionKling51-152 credits—3s, 4s, 5s, 6s, 7s, 8s
    Kling Video V3 Standard Motion ControlKling V3 Standard motion control - transfer motion from a driving video to a reference imageKling51-1210 credits720p—
    Kling Video V3 Standard Text to VideovariantKling V3 Standard text-to-video generation with camera controls and multi-prompt supportKling51-152 credits—3s, 4s, 5s, 6s, 7s, 8s

    12 February 2026

    6 models · 1 provider

    ByteDance shipped 6 video models on 12 February 2026. Of the 6, 2 have a spec page and 4 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 25 credits on Seedance 2.0 Fast. Highest output in it: 4k on Seedance 2.0.

    ModelBuilt byCredits per jobMax outputClip lengths
    Seedance 2.0Seedance 2.0 generates Hollywood-grade cinematic videos from text prompts with native audio-visual synchronization,…ByteDance44-3277 credits4k4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.0 FastvariantSeedance 2.0 Fast generates cinematic videos from text prompts with native audio-visual synchronization, director-level camera…ByteDance25-204 credits720p4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.0 Fast Image to VideovariantSeedance 2.0 Fast generates cinematic videos from reference images and text prompts with native audio-visual synchronization and…ByteDance55-204 credits720p4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.0 Fast Reference to VideoSeedance 2.0 Fast Reference to Video generates cinematic videos from reference images, videos, and audio inputs with native…ByteDance55-204 credits720p4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.0 Image to VideovariantSeedance 2.0 generates Hollywood-grade cinematic videos from reference images and text prompts with native audio-visual…ByteDance97-363 credits4k4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.0 Reference to VideovariantSeedance 2.0 Reference to Video generates cinematic videos from reference images, videos, and audio inputs with native…ByteDance97-363 credits4k4s, 5s, 6s, 7s, 8s, 9s

    5 March 2026

    5 models · 1 provider

    LTX shipped 5 video models on 5 March 2026. Of the 5, 3 have a spec page and 2 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 16 credits on LTX 2.3 Extend Video. Highest output in it: 4K on LTX 2.3 Image to Video Pro.

    ModelBuilt byCredits per jobMax outputClip lengths
    LTX 2.3 Extend VideovariantExtend an existing video — the model continues motion/action past the current end (or before the current start). Takes a source…LTX16-160 credits—2s, 4s, 6s, 8s, 10s, 12s
    LTX 2.3 Image to Video ProPro-quality image-to-video generation. Animates a still image with a prompt. Supports 1080p/1440p/2160p, 6-20s duration…LTX29-192 credits4K6s, 8s, 10s
    LTX 2.3 Retake VideoRetake a segment of an existing video. Takes a source video_url + prompt describing desired changes (e.g. "change flower to red…LTX16-160 credits—2s, 4s, 6s, 8s, 10s, 12s
    LTX 2.3 Text to Video FastFast text-to-video generation. Generates a video directly from a text prompt at 1080p/1440p/2160p, 6-20s duration, with optional…LTX20-256 credits4K6s, 8s, 10s, 12s, 14s, 16s
    LTX 2.3 Text to Video ProvariantPro-quality text-to-video generation. Generates a video directly from a text prompt at 1080p/1440p/2160p, 6-20s duration, with…LTX29-192 credits4K6s, 8s, 10s

    30 March 2026

    6 models · 1 provider

    Pixverse shipped 6 video models on 30 March 2026. Of the 6, 4 have a spec page and 2 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 10 credits on Pixverse V6 Image to Video. Highest output in it: 1080p on Pixverse Effects.

    ModelBuilt byCredits per jobMax outputClip lengths
    Pixverse EffectsAdd special effects to videos with PixversePixverse12-64 credits1080p5s, 8s, 10s
    Pixverse Image to VideovariantConvert images to videos with Pixverse V5.5Pixverse12-64 credits1080p5s, 8s, 10s
    Pixverse Text to VideoGenerate videos from text with Pixverse V5.5Pixverse12-64 credits1080p5s, 8s, 10s
    Pixverse TransitionCreate smooth transitions between video clipsPixverse12-64 credits1080p5s, 8s, 10s
    Pixverse V6 Image to VideovariantPixVerse V6 image-to-video. Animates a still image with strong motion and prompt control, optional audio and multi-clip, up to…Pixverse10-58 credits1080p5s, 8s
    Pixverse V6 Text to VideoPixVerse V6 text-to-video. Generates video from a text prompt with strong motion and prompt control, optional audio and…Pixverse10-58 credits1080p5s, 8s

    31 March – 6 April 2026

    5 models · 2 providers

    Google and Wan shipped 5 video models between 31 March 2026 and 6 April 2026. Of the 5, 4 have a spec page and 1 is folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 16 credits on VEO 3.1 Lite. Highest output in it: 4K on VEO 3.1 Lite.

    ModelBuilt byCredits per jobMax outputClip lengths
    VEO 3.1 LitevariantVEO 3.1 Lite is the most cost-effective model in the VEO 3.1 family for high-volume video generation, supporting text-to-video…Google16-52 credits4K4s, 6s, 8s
    Wan 2.7 Image to VideoWan 2.7 animates images with three modes: first-frame to video, first-and-last-frame interpolation, or video continuation, with…Wan16-180 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Wan 2.7 Reference to VideoWan 2.7 reference-to-video — generate videos from up to 5 reference images and/or videos, with optional first frame and voice…Wan20-144 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Wan 2.7 Text to VideoWan 2.7 generates high-fidelity videos from text prompts with strong motion consistency, optional custom audio input, and…Wan16-180 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Wan 2.7 Video EditWan 2.7 video editing — modify a source video using prompts and an optional reference image for character, clothing, or style…Wan20-144 credits1080p2s, 3s, 4s, 5s, 6s, 7s

    26 April 2026

    4 models · 1 provider

    Alibaba shipped 4 video models on 26 April 2026. Of the 4, 3 have a spec page and 1 is folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 34 credits on Happy Horse 1.0 Image to Video. Highest output in it: 1080p on Happy Horse 1.0 Image to Video.

    • New name on the roster: Alibaba.
    ModelBuilt byCredits per jobMax outputClip lengths
    Happy Horse 1.0 Image to VideovariantHappy Horse 1.0 animates a first-frame image into video with native synchronized audio, Foley sound effects, and multilingual…Alibaba34-336 credits1080p3s, 4s, 5s, 6s, 7s, 8s
    Happy Horse 1.0 Reference to VideoHappy Horse 1.0 generates videos from up to 9 reference images using character1–character9 placeholders in the prompt, with…Alibaba34-336 credits1080p3s, 4s, 5s, 6s, 7s, 8s
    Happy Horse 1.0 Text to VideoHappy Horse 1.0 generates expressive videos from text prompts with native synchronized audio, Foley sound effects, and…Alibaba34-336 credits1080p3s, 4s, 5s, 6s, 7s, 8s
    Happy Horse 1.0 Video EditHappy Horse 1.0 edits a source video using a prompt and up to 5 reference images, with auto/origin audio handling — output capped…Alibaba56-2688 credits1080p—

    1 May 2026

    3 models · 1 provider

    Grok shipped 3 video models on 1 May 2026. Of the 3, 1 has a spec page and 2 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 7 credits on Grok Imagine Video 1.5 Image to Video. Highest output in it: 1080p on Grok Imagine Video 1.5 Image to Video.

    ModelBuilt byCredits per jobMax outputClip lengths
    Grok Imagine Video 1.5 Image to VideovariantxAI's Grok Imagine Video 1.5 animates a still image with native audio. 480p, 720p or 1080p, any length from 1 to 15 seconds; the…Grok7-300 credits1080p1s, 2s, 3s, 4s, 5s, 6s
    Grok Imagine Video 1.5 Reference to VideovariantxAI's Grok Imagine Video 1.5 builds a video from up to 7 reference images, used as style and content guides and addressed in the…Grok7-168 credits720p1s, 2s, 3s, 4s, 5s, 6s
    Grok Imagine Video 1.5 Text to VideoxAI's Grok Imagine Video 1.5 turns a prompt into a video with native audio. 480p, 720p or 1080p, any length from 1 to 15 seconds,…Grok7-300 credits1080p1s, 2s, 3s, 4s, 5s, 6s

    19 May – 1 June 2026

    4 models · 2 providers

    ByteDance and Google shipped 4 video models between 19 May 2026 and 1 June 2026. Of the 4, 1 has a spec page and 3 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 22 credits on Seedance 2.0 Mini. Highest output in it: 4K on Gemini Omni Video.

    ModelBuilt byCredits per jobMax outputClip lengths
    Gemini Omni VideoGoogle Gemini Omni multimodal video generation. Accepts a prompt plus optional reference images, source video clips, character…Google29-72 credits4K4s, 6s, 8s, 10s
    Seedance 2.0 MinivariantSeedance 2.0 Mini is a faster, lower-cost tier of Seedance 2.0 — cinematic text-to-video with native audio-visual sync at high…ByteDance22-182 credits720p4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.0 Mini Image to VideovariantSeedance 2.0 Mini animates a first-frame image into cinematic video with native audio-visual sync — faster and cheaper than full…ByteDance49-182 credits720p4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.0 Mini Reference to VideovariantSeedance 2.0 Mini generates cinematic video from up to 9 reference images, 3 reference videos, and 3 reference audios…ByteDance49-182 credits720p4s, 5s, 6s, 7s, 8s, 9s

    10–17 June 2026

    3 models · 2 providers

    Bria and Kling shipped 3 video models between 10 June 2026 and 17 June 2026. Of the 3, 1 has a spec page, 1 is folded into a parent model's page as tier or mode variants and 1 has no page of its own. Cheapest complete job in the window: 2 credits on Bria Video Background Removal. Highest output in it: 1080p on Kling 3 Turbo Image to Video.

    • New name on the roster: Bria.
    ModelBuilt byCredits per jobMax outputClip lengths
    Bria Video Background RemovalAI-powered video background removal and replacementBria2 credits——
    Kling 3 Turbo Image to VideovariantKling 3 Turbo fast image-to-video generation.Kling22-108 credits1080p3s, 4s, 5s, 6s, 7s, 8s
    Kling 3 Turbo Text to VideoKling 3 Turbo fast text-to-video generation.Kling22-108 credits1080p3s, 4s, 5s, 6s, 7s, 8s

    23 June 2026

    3 models · 1 provider

    Alibaba shipped 3 video models on 23 June 2026. Of the 3, 1 has a spec page and 2 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 34 credits on Happy Horse 1.1 Image to Video. Highest output in it: 1080p on Happy Horse 1.1 Image to Video.

    ModelBuilt byCredits per jobMax outputClip lengths
    Happy Horse 1.1 Image to VideoHappy Horse 1.1 animates a first-frame image into 1080p video with synchronized native audio and multilingual lip-sync (aspect…Alibaba34-216 credits1080p3s, 4s, 5s, 6s, 7s, 8s
    Happy Horse 1.1 Reference to VideovariantHappy Horse 1.1 turns up to 9 reference images (character1–character9 placeholders) into 1080p video with synchronized native…Alibaba34-216 credits1080p3s, 4s, 5s, 6s, 7s, 8s
    Happy Horse 1.1 Text to VideovariantHappy Horse 1.1 is Alibaba's #1-ranked video model — generates 1080p video with synchronized native audio and multilingual…Alibaba34-216 credits1080p3s, 4s, 5s, 6s, 7s, 8s

    30 June 2026

    4 models · 1 provider

    Google shipped 4 video models on 30 June 2026. Of the 4, 1 has a spec page and 3 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 40 credits on Gemini Omni Flash Image to Video.

    ModelBuilt byCredits per jobMax outputClip lengths
    Gemini Omni Flash EditGoogle Gemini Omni Flash - conversational video editing (video-to-video)Google50-1200 credits——
    Gemini Omni Flash Image to VideovariantGoogle Gemini Omni Flash - animates a still image into video with audioGoogle40-100 credits—4s, 6s, 8s, 10s
    Gemini Omni Flash Reference to VideovariantGoogle Gemini Omni Flash - video with audio from multimodal reference imagesGoogle40-100 credits—4s, 6s, 8s, 10s
    Gemini Omni Flash Text to VideovariantGoogle Gemini Omni Flash - text-to-video with synchronized native audioGoogle40-100 credits—4s, 6s, 8s, 10s

    1 August 2026

    4 models · 2 providers

    ByteDance and NVIDIA shipped 4 video models on 1 August 2026. Of the 4, 2 have a spec page and 2 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 8 credits on Cosmos 3 Super Image to Video. Highest output in it: 720p on Cosmos 3 Super Image to Video.

    • New name on the roster: NVIDIA.
    ModelBuilt byCredits per jobMax outputClip lengths
    Cosmos 3 Super Image to VideoNVIDIA's Cosmos 3 Super world model animates a still image with physically grounded motion, guided by a text prompt that a…NVIDIA8-28 credits720p2s, 3s, 4s, 5s, 6s, 7s
    Seedance 2.5Dreamina Seedance 2.5 generates a native 30-second single-shot video at up to 720p from one text prompt, reasoning about the…ByteDance34-557 credits720p4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.5 Image to VideovariantAnimate a single still into a native 30-second clip at up to 720p, extending one frame into continuous motion without the drift…ByteDance34-557 credits720p4s, 5s, 6s, 7s, 8s, 9s
    Seedance 2.5 Reference to VideovariantGenerate video from up to 50 multimodal references — images, videos and audio — locking a character, set and palette across a…ByteDance34-557 credits720p4s, 5s, 6s, 7s, 8s, 9s

    5 August 2026

    7 models · 2 providers

    Black Forest Labs and MiniMax shipped 7 video models on 5 August 2026. Of the 7, 4 have a spec page and 3 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 34 credits on Flux 3 First Last Frame to Video. Highest output in it: 2K on Minimax H3 Image to Video.

    • New names on the roster: Black Forest Labs, MiniMax.
    ModelBuilt byCredits per jobMax outputClip lengths
    Flux 3 Extend VideoFLUX.3 continues an existing clip beyond its final frame, generating footage consistent with the original motion and scene…Black Forest Labs82-328 credits1080p5s, 6s, 7s, 8s, 9s, 10s
    Flux 3 First Last Frame to VideoFLUX.3 interpolates a smooth, coherent transition between a defined start frame and end frame, with native audio — up to 1080p…Black Forest Labs34-136 credits1080p5s, 6s, 7s, 8s, 9s, 10s
    Flux 3 Image to VideovariantFLUX.3 animates a single still image into coherent, natural motion with native audio — up to 1080p and 5–20 secondsBlack Forest Labs34-136 credits1080p5s, 6s, 7s, 8s, 9s, 10s
    Flux 3 Text to VideoFLUX.3 is Black Forest Labs' frontier video model — generates video with native audio directly from a text prompt, at up to 1080p…Black Forest Labs34-136 credits1080p5s, 6s, 7s, 8s, 9s, 10s
    Minimax H3 Image to VideovariantMiniMax H3 animates a supplied image into 2K video as the opening frame, or pairs a first and last frame to control a transition…MiniMax104-312 credits2K5s, 6s, 7s, 8s, 9s, 10s
    Minimax H3 Reference to VideovariantMiniMax H3 generates 2K video from multimodal references — up to 9 images for subject and style, 3 video clips for motion, and 3…MiniMax104-312 credits2K5s, 6s, 7s, 8s, 9s, 10s
    Minimax H3 Text to VideoMiniMax H3 is a frontier video model — generates 2K video from a text prompt alone, in durations from 5 to 15 seconds across six…MiniMax104-312 credits2K5s, 6s, 7s, 8s, 9s, 10s

    24–25 August 2026

    12 models · 3 providers

    LTX, MiniMax and Wan shipped 12 video models between 24 August 2026 and 25 August 2026. Of the 12, 3 have a spec page and 9 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 6 credits on Wan 3.0 Image to Video. Highest output in it: 2160p on LTX 2.5 Image to Video Fast.

    ModelBuilt byCredits per jobMax outputClip lengths
    LTX 2.5 Image to Video FastvariantSpeed-optimised image-to-video with synchronized native audio. Animates a still image, or interpolates between a start and end…LTX44-480 credits2160p6s, 8s, 10s, 12s, 14s, 16s
    LTX 2.5 Image to Video ProvariantQuality-optimised image-to-video with synchronized native audio, for final high-fidelity output. Animates a still image or…LTX58-136 credits1080p6s, 8s, 10s
    LTX 2.5 Text to Video FastvariantSpeed-optimised text-to-video with synchronized native audio in a single pass. 720p to 4K, 6-20s (durations past 10s render at…LTX44-480 credits2160p6s, 8s, 10s, 12s, 14s, 16s
    LTX 2.5 Text to Video ProQuality-optimised text-to-video with synchronized native audio in a single pass, for final high-fidelity output. 720p or 1080p,…LTX58-136 credits1080p6s, 8s, 10s
    Wan 3.0 Image to VideovariantAnimates a still image with Alibaba's Wan 3.0, with native audio. Ranked #1 on the Artificial Analysis image-to-video board at…Wan6-168 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Wan 3.0 Prime Image to VideovariantThe accelerated tier of Wan 3.0 image-to-video: animates a still image with native audio, served faster at a higher per-second…Wan11-336 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Wan 3.0 Prime Reference to VideovariantThe accelerated tier of Wan 3.0 reference-to-video: up to 10 images, 5 video clips and 5 audio clips, or a document or web page,…Wan11-336 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Wan 3.0 Prime Text to VideoThe accelerated tier of Wan 3.0: the same text-to-video model with native audio, served faster at a higher per-second rate. 480p,…Wan11-336 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Wan 3.0 Reference to VideovariantAlibaba's Wan 3.0 builds a video from reference media: up to 10 images, 5 video clips and 5 audio clips, addressed positionally…Wan6-168 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Wan 3.0 Text to VideoAlibaba's Wan 3.0 generates video with native audio from a text prompt. Ranked #1 on the Artificial Analysis text-to-video board…Wan6-168 credits1080p2s, 3s, 4s, 5s, 6s, 7s
    Minimax H3 Max Image to VideovariantAnimates a still image with a post-trained variant of MiniMax H3, tuned for stronger prompt adherence and better aesthetics. Add…MiniMax20-96 credits480P5s, 6s, 7s, 8s, 9s, 10s
    Minimax H3 Max Text to VideovariantA post-trained variant of MiniMax H3, tuned for stronger prompt adherence and better aesthetics, and roughly a third of H3's…MiniMax20-96 credits480P5s, 6s, 7s, 8s, 9s, 10s

    When 2026 was busy

    7 of the twelve months carried a video release; the busiest window was 1–5 February 2026, with 19.

    January 2026
    11
    February 2026
    25
    March 2026
    12
    April 2026
    8
    May 2026
    4
    June 2026
    13
    August 2026
    23

    Who shipped video in 2026

    Kling (20), LTX (14) and ByteDance (13) led on volume. SKUs, not quality — four tiers of one model count four times.

    ProviderModelsSpec pagesFirstLatest
    Kling2095 February 202617 June 2026
    LTX1496 January 202624 August 2026
    ByteDance1341 February 20261 August 2026
    Wan1066 April 202624 August 2026
    Pixverse8626 January 202630 March 2026
    Alibaba7426 April 202623 June 2026
    Google6331 March 202630 June 2026
    MiniMax515 August 202625 August 2026
    Black Forest Labs435 August 20265 August 2026
    Vidu4230 January 202630 January 2026
    Grok311 May 20261 May 2026
    Bria1010 June 202610 June 2026
    NVIDIA111 August 20261 August 2026

    What 2026 moved

    Firsts, measured against every video model released before them. “New name on the roster” is the provider label in the catalog, not the company — one lab can hold several labels.

    Dates are each model’s released_at value, walked in order and grouped until a window held 3 or more. Credits are what one complete generation costs per the model’s own price matrix; where it states none, the headline rate is shown and the model sits out the cheapest-in-window line.

    Other release years

    Frequently asked questions

    How many AI video models were released in 2026?+

    Versely's catalog carries 96 video models with a 2026 release date, from 13 providers, arriving in 16 launch windows between 6 January 2026 and 25 August 2026.

    What was the biggest AI video launch of 2026?+

    1–5 February 2026, with 19 video models from ByteDance and Kling. Of the 19, 9 have a spec page and 10 are folded into a parent model's page as tier or mode variants. Cheapest complete job in the window: 20 credits on DreamActor V2. Highest output in it: 4K on Kling Video V3 4K Image to Video.

    Which company released the most AI video models in 2026?+

    Kling, with 20 of the 96 video models dated 2026 — first on 5 February 2026, most recently on 17 June 2026. LTX shipped 14, ByteDance shipped 13, Wan shipped 10.

    What changed in AI video generation in 2026?+

    Measured against everything the catalog carried before it: New name on the roster: LTX; New name on the roster: Alibaba; New name on the roster: Bria.

    Run any 2026 video model in Versely

    The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync — in your browser or on your phone.