WeMM Video 2B

Language Tencent Partly verified
  • In Input modalities: TextImageAudioVideo Out Output modalities: Text
  • Reasoning
  • Open weights

Prices

0 offerings · prices as listed

No priced offering is listed for this model yet.

Details

Licence
apache-2.0
Family
wemm-video

Run locally

Weights: tencent/WeMM-Video-2B
  • Parameters 3.3B
  • Dense
  • Native precision BF16
  • Layers 24
  • KV cache per token 12 KB
  • Config positions 256K (config.json max_position_embeddings)
My hardware GeForce RTX 4090 24 GB · 1008 GB/s · 64 GB DDR5 dual channel · 89.6 GB/s
On your hardware · GeForce RTX 4090
Runs well BF16 (native) (automatic) 8K context 76–113 tok/sOne GPU
WeightsKV cacheOverheadUsable
Needs
≈ 6.7 GB (weights 6.1 (maker’s file) + KV cache 0.1 + overhead 0.5)
Usable
21.6 GB (90% of its memory)
Generation speed
76–113 tok/s
Longest context
256K (the model’s limit)
First token
about 0.11–0.16 s for a 2K prompt
PrecisionWeightsNeeds at 8KOn your hardwareLongest context
BF16 (native)maker’s file6.1 GB≈ 6.7Runs well 76–113 tok/s256K
Q8_03.2 GB≈ 3.8Runs well 140–211 tok/s256K
FP83.1 GB≈ 3.7Runs well 149–223 tok/s256K
Q6_K2.5 GB≈ 3.1Runs well 180–270 tok/s256K
Q5_K_M2.2 GB≈ 2.8Runs well 207–310 tok/s256K
Q5_02.1 GB≈ 2.7Runs well 214–321 tok/s256K
Q4_K_M1.9 GB≈ 2.5Runs well 241–361 tok/s256K
Q4_01.7 GB≈ 2.3Runs well 259–388 tok/s256K
IQ4_XS1.6 GB≈ 2.2Runs well 273–410 tok/s256K
Q3_K_Mquality drops1.5 GB≈ 2.1Runs well 296–443 tok/s256K
IQ3_XXSquality drops1.2 GB≈ 1.8Runs well 372–557 tok/s256K
Q2_Kquality drops1.1 GB≈ 1.7Runs well 378–568 tok/s256K
IQ2_XXSquality drops0.8 GB≈ 1.4Runs well 533–799 tok/s256K

Click a row to choose a precision. “Maker’s file” sizes are the maker’s own files; others are parameters × bits per weight.

Compared with other hardware

Each at the highest precision it runs. Click one for details.

Smallest that runs it (Q4_K_M, 8K): GPU GeForce RTX 3080; unified memory Apple M5 · 16 GB.

Run commands

vLLM
vllm serve tencent/WeMM-Video-2B --max-model-len 8192
Memory to fine-tune
Full≈ 51.6 GBDoesn’t fit your hardware
LoRA≈ 8.9 GBFits your hardware
QLoRA≈ 4.5 GBFits your hardware

Batch 1, 2,048-token sequences, gradient checkpointing; LoRA and QLoRA train rank-16 adapters on the attention projections.

Estimates; facts from the maker’s Hugging Face repositories, as of 2026-11-21. How it’s estimated

Cited scores

From third-party leaderboards, not measured by us

This model isn’t on any leaderboard we cite yet.

  1. New language model

Sources

Some values on this page are still waiting to be checked against official pages: a filled dot is verified, a hollow one is not yet (with or without a source on file). How we verify data

Tencent · China · tencent.com

Snapshots, the same family and other models from this provider