Wan Video

Text-to-video generation from Alibaba's Wan team, based on a flow-matching DiT with a UMT5-XXL text encoder and a causal 3D video VAE that streams decoding one latent frame at a time. mold implements the family natively in Rust.

• Developer: Wan-AI
• License: Apache 2.0
• Reference: Wan-Video/Wan2.1 · Wan-Video/Wan2.2

Note: Video output defaults to MP4. Also supports GIF, WebP, and APNG via --format. Frame count must be 4n+1 (77, 81, 121, ...) due to the VAE's 4x temporal compression. Wan generation currently targets CUDA; CPU runs are correctness-only.

Variants

Model	Steps	Approx total pull	Notes
wan21-t2v-1.3b:bf16	30	~14.5 GB	480p text-to-video; smallest, fastest pull
wan22-ti2v-5b:fp16	20	~22.8 GB	720p24 text- and image-to-video
wan22-t2v-a14b:q5	4	~36 GB	480p16 text-to-video, 4-step Lightning tier
wan22-t2v-a14b:q8	20	~42 GB	Same weights at Q8_0, no distill
wan22-i2v-a14b:q5	4	~36 GB	480p16 image-to-video, 4-step Lightning tier
wan22-i2v-a14b:q8	20	~42 GB	Same weights at Q8_0, no distill

Totals include the shared UMT5-XXL encoder (~11.4 GB), tokenizer, and the variant's VAE. The encoder is shared across every Wan model under shared/wan/, so a second Wan pull only fetches the checkpoint and VAE.

A14B is two models

Wan 2.2 A14B is a mixture of experts along the noise axis: two complete 14B transformers, one trained for the early, structural part of the schedule and one for the late, detail part. mold loads the high-noise expert first, switches once when the schedule crosses the boundary (timestep 875 for T2V, 900 for I2V), and drops each expert before loading its partner — so VRAM is the larger of the two experts, not their sum (~10.8 GB at :q5, ~15.4 GB at :q8). Disk is the sum, which is why the pull totals are large.

The :q5 tier additionally pulls lightx2v's 4-step distill — a separate adapter for each expert — and defaults to guidance 1.0. That is not a weak setting: at guidance ≤ 1 mold skips the unconditional pass entirely, so each step is one forward instead of two. Four steps at one forward each is where the tier's speed comes from.

Usage

# 480p, 81 frames @ 16 fps (defaults)
mold run wan21-t2v-1.3b "a red fox trotting through fresh snow, golden hour"

# 720p24, 121 frames — Wan 2.2 5B
mold run wan22-ti2v-5b "aerial view of waves breaking on a black sand beach" \
  --width 1280 --height 704 --frames 121 --fps 24

# Wan 2.2 A14B, 4-step Lightning tier
mold run wan22-t2v-a14b:q5 "a paper boat drifting down a rain gutter"

# A14B image-to-video from a still
mold run wan22-i2v-a14b:q5 "the balloon lifts off" --image balloon.png

Wan checkpoints were tuned against a specific negative prompt; mold applies it automatically when --negative is not given.

Defaults and limits

Property	wan21-t2v-1.3b	wan22-ti2v-5b	wan22-*-a14b:q5	wan22-*-a14b:q8
Resolution	832x480 / 480x832	1280x704 / 704x1280	832x480	832x480
Frames	81 @ 16 fps	121 @ 24 fps	81 @ 16 fps	81 @ 16 fps
Steps	30	20	4	20
Guidance	6.0	5.0	1.0 (no CFG pass)	3.5
Flow shift	8.0	8.0	5.0	5.0
Sampler	FlowUniPC (bh2)	FlowUniPC (bh2)	FlowUniPC (bh2)	FlowUniPC (bh2)

The sampler schedule matches the one lightx2v's Lightning distills were trained against (diffusers' flow-UniPC grid), so the 4-step tier reproduces its published timesteps exactly. Override the flow shift with MOLD_WAN_SHIFT.

Quantized checkpoints and adapters

A14B ships as GGUF. Quantized weights stay quantized in memory and dequantize inside the matmul, which is what keeps a 14B expert at ~10.8 GB rather than ~28 GB. A LoRA cannot be merged into a weight in that state without requantizing it, so on GGUF mold applies adapters as a parallel low-rank branch instead — the same arithmetic, applied at full precision, with no load cost. On bf16 and fp8-scaled safetensors the adapter is merged as the weights are read.

*_fp8_e4m3fn_scaled safetensors also load: the weights stay 1 byte per parameter and dequantize per call against their per-module scale. The e5m2 variants some repositories publish beside them are refused by name — mold reads the e4m3 flavour only.

Discovery

The models in the table above install by name. Community Wan fine-tunes are additionally discoverable in the catalog — open Models → Discover in Mold Studio and search for wan, then install the row, or pull a Civitai version id directly:

mold pull wan22-t2v-a14b:q5   # manifest name — the A14B fast tier
mold pull wan21-t2v-1.3b      # bare names resolve their default tag
mold pull cv:<version-id>     # a catalog row

mold pull takes manifest names and catalog ids (cv:…, hf:…). A catalog id has to name a row the catalog actually supports, which is not the same as any Hugging Face repository — hf:Wan-AI/Wan2.2-T2V-A14B, for instance, is the upstream A14B pair and is not installable for the reason below. When in doubt, install from Models → Discover, which only lists rows this build can run.

Every Wan checkpoint in the wild ships the transformer alone, so a catalog install also pulls the shared UMT5-XXL encoder and the matching VAE. Those are the same files the manifest models use, under shared/wan/, so a second Wan install reuses them.

What the catalog deliberately does not offer:

• Wan 2.1 image-to-video conditions through a CLIP-vision cross-attention branch mold's transformer does not implement, so the download would install and then fail to generate.
• Wan 2.5 and 2.7 are later architectures with no mold engine.
• A14B is a two-expert pair, and a catalog row installs one file. Civitai publishes the high- and low-noise experts as separate versions, so an installed row would denoise the whole schedule with whichever expert was pulled. Use the manifest tiers above, which ship both experts.

Wan 2.1 text-to-video at either size and TI2V-5B install from the catalog normally.

Roadmap

Remaining Wan work is tracked in the Wan Video milestone.