AOpen

Wan 2.2

Alibaba (Tongyi Lab)

MoE architecture with 27B total params but only 14B active. Trained on 65% more images and 83% more video than 2.1. Outperforms leading closed-source models on Wan-Bench 2.0.

AOpen

SkyReels V2

Skywork AI

First open model for infinite-length video. Best human rendering in open source.

BOpen

HunyuanVideo 1.5

Tencent

Punches above its weight. Avatar and Custom variants add versatility.

Pick Wan 2.2 if…

You want cinematic style control, speech-to-video, or consumer GPU deployment (TI2V-5B).

Pick SkyReels V2 if…

You want long-form AI films, human-centric content, or short films and ads.

Pick HunyuanVideo 1.5 if…

You want avatar gen, custom characters, or research.

Specifications

Maker
Alibaba (Tongyi Lab)
Skywork AI
Tencent
Source Type
Open Source
Open Source
Open Source
License
Apache 2.0
Open Source
Tencent Open (check terms)
Architecture
DiT + MoE (2-expert: high-noise + low-noise)
AR Diffusion-Forcing
DiT + 3D Causal VAE
Parameters
27B total (14B active per step, 2x14B experts)
14B
8.3B
Max Resolution
720p
720p
720p
Max Duration
10-15s
30s+ (infinite)
6-10s
FPS
24
24
24
Native Audio
No
No
No
ComfyUI Support
Yes
No
Yes
Fine-tunable
Yes
Yes
Yes
Min VRAM
8GB (small) / 24GB (full)
40GB+ (A100/H100)
24GB (RTX 4090)
Cost / Second
Self-host
Self-host
$0.06
Inputs
T2V (A14B), I2V (A14B), TI2V (5B), S2V (14B)
T2V, I2V
T2V, I2V, Avatar
On Floyo
Yes
No
Yes

Strengths & Trade-offs

Wan 2.2

Strengths

  • +First MoE in video diffusion
  • +27B total but only 14B active per step
  • +high-noise expert for layout + low-noise for detail
  • ++65.6% more images and +83.2% more video training data vs 2.1
  • +cinematic aesthetic control (lighting, composition, contrast, color tone)

Trade-offs

  • -720p cap
  • -MoE needs careful threshold tuning (SNR-based)
  • -no native audio in base model (S2V is separate)
  • -newer ecosystem than 2.1

Best For

  • Self-hosted production
  • cinematic style control
  • speech-to-video
  • consumer GPU deployment (TI2V-5B)

SkyReels V2

Strengths

  • +First infinite-length open model
  • +10M+ film/TV training
  • +best open human faces

Trade-offs

  • -Heavy GPU req
  • -no ComfyUI
  • -newer community
  • -720p cap

Best For

  • Long-form AI films
  • human-centric content
  • short films and ads

HunyuanVideo 1.5

Strengths

  • +Efficient 8.3B params
  • +~75s on 4090
  • +Avatar + Custom variants

Trade-offs

  • -Short output
  • -720p max
  • -no audio

Best For

  • Avatar gen
  • custom characters
  • research
  • consumer GPU