What GPU do I need to run Wan-AI/Wan2.1-T2V-1.3B?
1.4B parameters, published in F32. View on Hugging Face
Set up Wan2.1-T2V-1.3BWan2.1-T2V-1.3B is published by Wan-AI on Hugging Face, with 22,818 downloads and 522 likes to date. It's a unlisted-architecture model built for text-to-video, published natively in F32.
VRAM required & cheapest live GPU fit
Required VRAM = weight size at each precision, plus a fixed overhead for activation memory and allocator fragmentation. Diffusion and video models carry no KV-cache. The real driver of extra memory is output resolution and frame count, which this flat overhead does not model. Full formula and assumptions: methodology.
| Precision | Weight size | Required VRAM | Cheapest live fit | GPUs needed | Est. $/hr (full fit) |
|---|---|---|---|---|---|
| FP32 | 5.3 GB | 6.3 GB | V100 | 1 | $0.088/hr |
| FP8 (quantized) | 1.3 GB | 1.6 GB | RTX 4070 Super | 1 | $0.121/hr |
| INT4 (quantized) | 0.7 GB | 0.8 GB | RTX 3060 | 1 | $0.110/hr |
A GPU is only matched to a row if its hardware supports that precision, and the primary recommendation is always a single-GPU fit when one exists.
INT4 requires a quantized checkpoint actually published for this model, check its Hugging Face page before relying on this row.
Cheapest way to run Wan2.1-T2V-1.3B at its published (F32) precision: 1× V100, at $0.088/hr per GPU ($0.088/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
Wan2.1-T2V-1.3B: KV cache by context length
The KV cache is the memory the attention layers hold for every token of context, on top of the weights. It is not part of the flat 1.2x overhead in the table above, grows with context length and with every concurrent request, and is why a long-context deployment needs more VRAM than the table shows.
Multi-head attention: every layer caches keys and values for every head. Cached per token: 184,320 bytes at 16-bit.
The model's maximum context length is not published in its configuration, so no per-context figures are shown.
Computed from the layer, head and window counts in the model's own configuration (Wan-AI/Wan2.1-T2V-1.3B), at 16 bits per cached value. Each concurrent sequence needs its own cache, on top of the weights.
Wan2.1-T2V-1.3B: common questions
Does Wan2.1-T2V-1.3B fit on a 8 GB GPU?
Yes. At FP32 it needs 6.3 GB of VRAM, so an 8 GB card holds it with 1.7 GB to spare. A 6 GB card is not enough for it at FP32.
Can Wan2.1-T2V-1.3B run in 16-bit instead of FP32?
Yes. Its published weights are FP32, 5.3 GB, or 6.3 GB once inference overhead is added. Loading the same weights in 16 bits halves that to 2.6 GB, or 3.2 GB with overhead. That moves it onto a 6 GB card instead of an 8 GB one. How much accuracy the cast costs is model-specific and is not measured here.
How many copies of Wan2.1-T2V-1.3B fit on one V100?
2, by VRAM alone. That card carries 16.0 GB and one copy needs 6.3 GB at FP32, on a live rate of $0.088/hr for the whole card. Throughput is not modelled here, so 2 copies is not 2 times the requests served.
What is the least VRAM Wan2.1-T2V-1.3B can run in?
0.8 GB, at INT4 (quantized), which fits a 6 GB card, against 6.3 GB at FP32. That row assumes an INT4 checkpoint has actually been published for this model, so check its Hugging Face page before planning around it.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
More Wan2.1 models
- Wan2.1-T2V-1.3B-Diffusers (1.4B, F32)
- TurboWan2.1-T2V-1.3B-Diffusers (1.4B, BF16)
- Wan2.1-VACE-1.3B-diffusers (2.2B, F32)
- Wan2.1-T2V-14B-Diffusers (14.3B, F32)
- Wan2.1-T2V-14B (14.3B, F32)
Alternatives at this size
Other models for text-to-video within about a third of Wan2.1-T2V-1.3B's 1.4B parameters, from other model lines.
- CogVideoX-2b (1.7B, F16)
- text-to-video-ms-1.7b (1.4B, F32)
- LTX-Video-0.9.5 (1.9B, BF16)
- i2vgen-xl (1.4B, F32)
More on Wan2.1-T2V-1.3B
Fits on an 8 GB GPU at FP32: every model that fits in 8 GB.
Best video generation models: how Wan2.1-T2V-1.3B ranks against the rest.
Related reading: V100 pricing and specs, Fast diffusion inference on GPU VMs, and What AI inference is.
