tencent · video

HunyuanVideo-I2V

tencent/HunyuanVideo-I2V

HunyuanVideo-I2V at Q4_K_M is exactly 7,883,680,512 bytes (7.34 GiB / 7.88 GB) — an effective 4.919 bits per weight, not the nominal 4.

From the file· summed from 1 file(s)
Parameters
12.8B
Architecture
hyvid
Context
—
native (config.json)
License
other

Shipped quantizations

● exact bytes, summed from published files
QuantSize●Exact bytes●Effective bpw●Tensors●Publisher
Q2_K2.96 GiB3,179,434,3041.984—calcuis
Q3_K_S5.67 GiB6,087,666,4323.799—city96
Q3_K_M5.81 GiB6,238,071,5523.892—city96
Q4_K_S7.21 GiB7,742,122,7524.831—city96
Q4_07.21 GiB7,742,122,7524.831—city96
Q4_K_M7.34 GiB7,883,680,5124.919—city96
Q4_17.94 GiB8,520,690,4325.317—city96
Q5_08.66 GiB9,299,258,1125.803—city96
Q5_K_S8.66 GiB9,299,258,1125.803—city96
Q5_K_M8.80 GiB9,449,663,2325.896—city96
Q5_19.39 GiB10,077,825,7926.288—city96
Q6_K10.20 GiB10,953,714,4326.835—city96
Q3_K_M2 shards11.32 GiB12,155,696,4167.585—calcuis
Q8_013.01 GiB13,970,664,1928.717—city96
Q4_K_M2 shards14.40 GiB15,464,609,0569.649—calcuis
Q4_02 shards14.40 GiB15,464,609,0569.649—calcuis
Q4_12 shards15.85 GiB17,021,744,41610.621—calcuis
Q5_02 shards17.30 GiB18,578,879,77611.593—calcuis
Q5_K_M2 shards17.44 GiB18,729,284,89611.687—calcuis
Q5_12 shards18.75 GiB20,136,015,13612.564—calcuis
Q6_K2 shards20.38 GiB21,887,792,41613.658—calcuis
BF1623.89 GiB25,649,179,39216.005—city96
Q8_02 shards26.00 GiB27,921,691,93617.422—calcuis
BF162 shards49.08 GiB52,699,804,96032.883—calcuis

Compare with

same modality, comparable size
vs LTX-2.3vs Wan2.2-I2V-A14Bvs Wan2.2-T2V-A14Bvs Wan2.2-Animate-14Bvs Sulphur-2-basevs WAN2.2-14B-Rapid-AllInOneAll alternatives →Will it fit on my card? →

Will it run on your card?

full quant x context sweep
By VRAM:8GB12GB16GB24GB32GB48GB
Radeon RX 6500 XT 4GBGeForce RTX 3050 6GBGeForce RTX 5050 8GBGeForce RTX 3080 10GBGeForce RTX 2080 Ti 11GBGeForce RTX 5070 12GBGeForce RTX 5060 Ti 16GBApple M3 Pro 18GBGeForce RTX 3080 Ti 20GBGeForce RTX 5090 D V2 24GBGeForce RTX 5090 D 32GBApple M5 Max 36GBApple M5 Pro 48GBApple M5 Max 64GBApple M3 Ultra 96GBApple M5 Max 128GBApple M2 Ultra 192GBApple M3 Ultra 256GBApple M3 Ultra 512GB

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_K_M at roughly 6.72 GiB. The real file is 7.34 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

Architecture unavailable — this repository is gated and no ungated mirror was found. Exact file sizes above are still authoritative; only the KV math needs the config.

Questions people ask

How much VRAM does HunyuanVideo-I2V need?
Q4_K_M is exactly 7,883,680,512 bytes (7.34 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of HunyuanVideo-I2V should I use?
Q4_K_M is the usual default. Pick the largest quantization that fits your card at the context you actually need — the table above gives exact sizes for every one published.