Wan 2.2 5B · 32 cards measured first-party · Updated July 2026

How Fast Is Wan 2.2 at 720p on Each GPU?

Wan 2.2 TI2V-5B generates a 49-frame clip at 1280×704, real 720p AI video. It needs ~18GB of VRAM minimum and ~38GB of system RAM, and it is the most honest test in our suite of whether a machine can do modern video generation or just talk about it.

Benchmarked weights: Wan-AI/Wan2.2-TI2V-5B-Diffusers

Fastest we measured

NVIDIA B300

2.94 frames/s on Wan 2.2 5B. Measured on our bench. 288GB of VRAM, 1400W board rating. AI Score 93.8/100 across our full 12-workload suite.

Pros

• 2.94 frames/s on Wan 2.2 5B
• 288GB, clears the Wan 2.2 5B floor
• Rentable by the hour rather than bought

Cons

• 1400W board rating
• Datacenter or workstation hardware, not a retail purchase

Best for: Wan 2.2 5B work where you want the ceiling gone rather than the cheapest entry.

Runner-up

NVIDIA B200

1.88 frames/s on Wan 2.2 5B. Measured on our bench. 192GB of VRAM, 1000W board rating. AI Score 78.0/100 across our full 12-workload suite.

Pros

• 1.88 frames/s on Wan 2.2 5B
• 192GB, clears the Wan 2.2 5B floor
• Rentable by the hour rather than bought

Cons

• 1000W board rating
• Datacenter or workstation hardware, not a retail purchase

Best for: Wan 2.2 5B work where you want the ceiling gone rather than the cheapest entry.

Third

NVIDIA H200

1.41 frames/s on Wan 2.2 5B. Measured on our bench. 141GB of VRAM, 700W board rating. AI Score 65.0/100 across our full 12-workload suite.

Pros

• 1.41 frames/s on Wan 2.2 5B
• 141GB, clears the Wan 2.2 5B floor
• Rentable by the hour rather than bought

Cons

• 700W board rating
• Datacenter or workstation hardware, not a retail purchase

Best for: Wan 2.2 5B work where you want the ceiling gone rather than the cheapest entry.

Compute-bound and heavy. On our measured B300 this workload ran at 95.9% utilisation, drew over a kilowatt, and produced the hottest temperature we logged across all 21 workloads. Video is the one thing that reliably makes a datacenter GPU behave like a datacenter GPU.

Because this workload is tensor-compute bound, the ranking tracks architecture generation and tensor throughput rather than memory bandwidth, the reverse of our LLM charts. The same two cards can swap places entirely depending on which of these pages you're reading. That's the reason we run twelve workloads instead of publishing one score. A GPU isn't fast or slow. It's fast at some things and gated out of others, and which of those matters depends entirely on what you're actually going to run.

Our verdict

NVIDIA B300 tops our Wan 2.2 5B leaderboard at 2.94 frames/s (measured), 2940% of the way clear of the slowest card that still fits. But the number that decides most purchases isn't on the chart. It's the 22 cards that can't run Wan 2.2 5B at all. This is a compute workload: buy architecture generation, not raw VRAM, as long as you clear the floor first.

FAQ

What is the fastest GPU for Wan 2.2 5B?

NVIDIA B300, at 2.94 frames/s on our bench, a first-party measurement. It carries 288GB of VRAM. Of the 54 cards we have Wan 2.2 5B data for, 32 can run it at all.

How much VRAM do I need for Wan 2.2 5B?

~18GB VRAM minimum, ~34GB for the full path, plus ~38GB of system RAM. The RAM requirement gates as many machines as the VRAM does.

Why does the Wan 2.2 5B ranking look different from your other benchmarks?

Because this workload is tensor-compute bound, the ranking tracks architecture generation and tensor throughput rather than memory bandwidth, the reverse of our LLM charts. The same two cards can swap places entirely depending on which of these pages you're reading. That's why we publish twelve separate workloads rather than one blended score, the ordering genuinely changes depending on the job.

Are these Wan 2.2 5B numbers measured or estimated?

Both, and every row says which. 32 of the 54 cards here were rented and run by us on the same harness. The remainder are anchored estimates interpolated per workload against those measurements. We never blend the two silently, if a row says Estimated, we have not run that card.

Can I rent a GPU to run Wan 2.2 5B instead of buying one?

Yes, and for the cards at the top of this leaderboard it's the only realistic option, most of them have no retail channel at all. It's also how we got these numbers: we rented the hardware by the hour rather than buying it. That's worth considering before you spend on a card to find out whether it's fast enough.

Why publish cards that can't run Wan 2.2 5B?

Because it's the most useful thing we know. A card that can't load a model doesn't run it slowly, it doesn't run it. Most benchmark sites leave that as a blank cell or quietly drop to a smaller quantisation to produce a number. We publish it as a hard gate and score it zero, because 'this card cannot do the thing you want' is the answer to the question you were actually asking.

How we test

Every ranking on this page comes from our own benchmark runs, not vendor claims. Cards marked Measured were rented and run by us; cards marked Estimated are interpolated per workload against those measured anchors and are labelled on every row, we never blend the two silently. LLMs run on llama.cpp (llama-bench) at Q4_K_M with -p 512 -n 128. Diffusion and video run on diffusers/ComfyUI at BF16, with SDXL at FP16. Each workload gets a warmup pass plus multiple timed runs (5 for small LLMs, 3 for large models and images, 2 for video); we publish the mean as the result and the minimum as the 1% low. Run-to-run variance is under 0.5%. Telemetry, power, temperature, utilisation, clocks, peak VRAM, is sampled at 1 Hz for the duration of every run. Where a model exceeds a card's VRAM we publish a hard won't-fit result rather than quietly dropping to a smaller quantisation. A card that can't run a model scores zero on it. Silently swapping precision to make a number appear would make every number on this site meaningless. All figures are single-GPU, single-stream, batch-size-1. That is the honest way to measure what one card does for one user, and it is deliberately not how a datacenter serves a model. Vendor and MLPerf figures use large batches across many GPUs and will be far higher. Neither is wrong, they answer different questions. Ours answers 'what will this card do for me'.