SkyReels
SkyworkSkyReels is Skywork AI's line of video generation models. The current release, SkyReels V4, described in February 2026, generates video and synchronized audio together from text, image, video and audio prompts at up to 1080p and 15 seconds, and also handles inpainting and editing. Earlier generations, through SkyReels V3, were published with weights.
SkyReels V4 has no published weights. SkyReels V3 (January 2026; reference-to-video, video extension and talking-avatar models) and V2 (April 2025) remain downloadable under the Skywork Community License. The SkyReels-A1 and A2 portrait-animation models are a separate series and are not part of this entry.
Openness
1 medium confidence- weights
- closed (no SkyReels V4 weights on Skywork's Hugging Face account, where a search for V4 returns nothing, or in its GitHub organization, which has no V4 repository)
- data
- closed (nothing is published for SkyReels V4 beyond its technical report)
- code
- closed (no code is published for SkyReels V4)
- license
- Proprietary (no license is published for SkyReels V4, which is available only as a hosted model)
- prior-release
- distributed (SkyReels V3 and V2: checkpoints download from the Hub without a gate under the Skywork Community License, which permits commercial use; inference code only)
SkyReels V4, Skywork's current release, has no published weights; Skywork describes it in a technical report and it is ranked on public arenas as a hosted model. SkyReels V3 and V2 remain downloadable under the Skywork Community License, which allows commercial use, so the open line trails the vendor's newest release.
- https://artificialanalysis.ai/video/leaderboard/text-to-video recorded 2026-11-12
"SkyReels V4" by Skywork AI is listed with "openWeightsUrl":null, a release date of 2026-03-18 and a per-minute API price.
- https://arxiv.org/abs/2602.21818 recorded 2026-11-12
Skywork's SkyReels-V4 technical report, "[Submitted on 25 Feb 2026 (v1), last revised 18 Mar 2026]": a unified model for joint video-audio generation, inpainting and editing, "up to 1080p resolution, 32 FPS, and 15 second duration"; no code, weights or training-data link.
- https://huggingface.co/api/models?author=Skywork&search=SkyReels&sort=downloads&direction=-1&limit=100 recorded 2026-11-12
Listing of the Skywork SkyReels repositories: V1, V2, V3, A1 and A2 checkpoints, the newest the SkyReels-V3 models of 2026-01-19; no V4 repository.
- https://huggingface.co/api/models?author=Skywork&search=V4 recorded 2026-11-12
Hub search of the Skywork account for "V4" returns an empty list.
- https://huggingface.co/api/models/Skywork/SkyReels-V3-A2V-19B?expand[]=downloads&expand[]=likes&expand[]=cardData&expand[]=gated&expand[]=lastModified&expand[]=createdAt&expand[]=siblings recorded 2026-11-12
Hugging Face API record for Skywork/SkyReels-V3-A2V-19B: license_name skywork-license, gated false, safetensors shards.
- https://huggingface.co/Skywork/SkyReels-V3-A2V-19B/raw/main/LICENSE recorded 2026-11-12
SkyReels V3 LICENSE: "The community usage of Skywork model requires Skywork Community License. The Skywork model supports commercial use."
- https://ungh.cc/orgs/SkyworkAI/repos recorded 2026-11-12
SkyworkAI GitHub organization repositories: SkyReels-V1, SkyReels-V2, SkyReels-V3, SkyReels-A1, SkyReels-A2; no SkyReels-V4 repository.
- https://ungh.cc/repos/SkyworkAI/SkyReels-V3/files/main/README.md recorded 2026-11-12
SkyReels-V3 README: "Jan 29, 2026: We release the inference code and model weights of SkyReels-V3"; the same day an API was launched.
Adoption
2 high confidenceAdoption is measured as Hugging Face downloads summed across the Skywork SkyReels V1 to V3 checkpoints, most of them V2. The current V4 has no downloadable weights to count.
- https://huggingface.co/api/models?author=Skywork&search=SkyReels&sort=downloads&direction=-1&limit=100 recorded 2026-11-12
Listing sorted by downloads: SkyReels-V2-DF-14B-720P 7836, SkyReels-V2-T2V-14B-720P 7769, SkyReels-V1-Hunyuan-T2V 4961, SkyReels-V3-A2V-19B 2558, the rest under 500 each; 26,164 in total, of which the A1 and A2 repositories account for 80.
Capability
3 medium confidenceSkyReels V4 places in the upper half of the public text-to-video arena, behind MiniMax H3, Wan 3.0 and Seedance, and higher on the image-to-video board. The open V3 and V2 models are not ranked there.
- https://artificialanalysis.ai/video/leaderboard/image-to-video recorded 2026-11-12
Without-audio image-to-video list (40 models): "SkyReels V4" 6th, Elo 1120.
- https://artificialanalysis.ai/video/leaderboard/text-to-video recorded 2026-11-12
Without-audio list (39 models): "SkyReels V4" 11th, Elo 1104; "MiniMax H3 (768p)" 6th with an open-weights link; "MiniMax H3 Max" 2nd; "Wan 3.0" 4th.
Verified 2026-11-12, except some scores, still awaiting re-confirmation