selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

tencent logo

WeVisCut

tencent·Instruct·released 9 Sept 2026

Custom licence9.4B256K ctx ≈7.1 GB
Your GPU
llama-server -hf prithivMLmods/tencent-WeVisCut-GGUF:Q4_K_M
All 3 runners ↓

WeVisCut is a 9B tool-using agent that turns knowledge articles or narration scripts into multi-minute edited videos. At each step, it reads the current editing state and selects one of 12 tools for planning, narration, asset retrieval, composition, inspection, or repair.

Notify me when

How much VRAM does WeVisCut need?

At Q4_K_M WeVisCut needs about 7.1 GB with an 8K context (7.1 GB at Q4, 20 GB at full precision). Pick your GPU to see what fits and how fast it runs.

--vram
any
--gpus
--quant

Needs: ≈7.0 GB · Q4_K_M · 8K context — Pick a GPU to check the fit.

Change quant, context or KV cache to compare
Advanced: context 8K · KV cache 16-bit
Context
KV cache

Smallest GPU for each quant and context (KV cache 16-bit)

Quantisation8K16K32K64K128K256K
Q3_K_M
Q4_K_M
Q5_K_M
Q6_K
Q8_0
FP16

≤16 GB card24–32 GB48–96 GBmore

QuantisationQualityWeightsKV cacheTotal
Q2_KHeavy lossSmallest, noticeably worse qualityHeavy loss2.9 GB est.1.0 GB≈4.5 GB
Q3_K_MNoticeable lossSmall, some quality lossNoticeable loss4.3 GB file1.0 GB≈6.0 GB
Q4_K_MrecommendedSmall lossRecommended default for local useSmall loss5.2 GB file1.0 GB≈7.0 GB
Q5_K_MNear-losslessNear-lossless for most tasksNear-lossless6.0 GB file1.0 GB≈7.8 GB
Q6_KNear-losslessPractically losslessNear-lossless6.8 GB file1.0 GB≈8.7 GB
Q8_0LosslessLossless in practice, twice the size of Q4Lossless8.9 GB file1.0 GB≈11 GB
FP16 / BF16LosslessFull precision, as releasedLossless17 GB file1.0 GB≈19 GB

Estimates for 8K tokens of context: weights + KV cache + runtime overhead. Speed assumes memory-bound decoding at your GPU’s bandwidth. Pick a GPU above and the fit column checks every row against it. Weights marked “file” use the real GGUF download size; “est.” rows are computed from the parameter count.

How to pick a quant

Take the biggest quant that fits with room left for context. Q4_K_M is the usual sweet spot. Below Q4 quality drops fast, above Q6 you mostly pay memory for nothing. Big models (70B+, large MoE) tolerate low quants better than small ones.

  • LosslessF16, Q8_0
  • Near-losslessQ6_K, Q5_K_M
  • Small lossQ5_K_S, Q4_K_M, IQ4_NL, IQ4_XS, Q4_K_S
  • Noticeable lossQ4_0, Q3_K_L, Q3_K_M, IQ3_M
  • Heavy lossIQ3_XXS, Q3_K_S, IQ2_M, Q2_K

Typical quality loss, not measured per model. Tiers come from KL divergence against FP16 on Llama 3 8B: llama.cpp perplexity README: LLaMA 3 8B scoreboard (KLD and ΔPPL vs FP16, Wikitext-2); llama.cpp quantize tool: ΔPPL per quant type @ Llama-3-8B

How much context fits?

Model quant
KV cache

At 8K context, Q4_K_M needs ≈7.0 GB(5.2 GB weights + 1.0 GB KV cache + 0.8 GB overhead). Pick a GPU above to see how much context fits.

0102030401K2K4K8K16K32K64K128K256KGB

Weights + overheadKV cache

Can the RTX 4060 run WeVisCut? →

Speed vs quality per quant

Smaller quants run faster and fit in less memory, but lose quality.

Q2_KQ3_K_MQ4_K_MQ5_K_MQ6_KQ8_0F16

Fits in VRAMTight fitCPU offload✕ won't runQuality is typical KL divergence vs FP16, not measured for this model.

Real-world speed of WeVisCut

Tokens per second measured by people running it at home. A GPU shows a median once 3 reports agree on the same setup. Reports far from our estimate wait for a review. Measure yours with llama-bench →

No medians yet. Be one of the first 3 to report a setup.

All reports

Run WeVisCut locally

--quantQ4_K_Mdefault; pick your GPU to tailor itprithivMLmods/tencent-WeVisCut-GGUF

OpenAI-compatible server with a web UIdocs
llama-server -hf prithivMLmods/tencent-WeVisCut-GGUF:Q4_K_M

Downloads the GGUF on first run. Installed through llama.app? The same flags work with `llama serve`.

Install llama.cpp

Serve an OpenAI-compatible APIdocs
vllm serve tencent/WeVisCut
Docker Compose (NVIDIA)docs
yaml
services:
  vllm:
    image: vllm/vllm-openai:latest
    command: tencent/WeVisCut
    ports:
      - "8000:8000"
    ipc: host
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

API on port 8000.

Install vLLM

Serve an OpenAI-compatible APIdocs
sglang serve tencent/WeVisCut --host 0.0.0.0 --port 30000
Docker Compose (NVIDIA)docs
yaml
services:
  sglang:
    image: lmsysorg/sglang:latest
    command: sglang serve tencent/WeVisCut --host 0.0.0.0 --port 30000
    ports:
      - "30000:30000"
    ipc: host
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

API on port 30000.

Install SGLang

  • Transformersofficial · library

Quantised versions and downloads

About WeVisCut

WeVisCut is a 9B tool-using agent that turns knowledge articles or narration scripts into multi-minute edited videos. At each step, it reads the current editing state and selects one of 12 tools for planning, narration, asset retrieval, composition, inspection, or repair. It is not a standalone text-to-video model; use it with the WeVisCut runtime, which executes the model's tool calls. Requirements: Python 3.11 or later and an NVIDIA GPU with about 24 GB of memory for bf16 inference. Install a PyTorch build compatible with your CUDA version before the GPU requirements below. The setup script installs Node.js, FFmpeg, and the renderer. Open http:// :8081/ to use the Studio. For UI-only mode without loading model weights, run bash run_studio.sh --no-model. The policy is based on Qwen3.5-9B and is trained on executable editing trajectories. Its runtime provides an editable timeline with pause/resume, manual and chat-based revisions, per-shot rerendering, and version rollback. Video quality depends on the configured retrieval, TTS, and rendering services and available templates. WeVisCut is intended for research and production assistance in long-form video editing. It does not guarantee factual accuracy or production quality; review generated narration and video before publication. The model weights require the WeVisCut tool environment to produce a video.

Full model card on Hugging Face →

Alternatives to WeVisCut

All alternatives →
google logo

gemma-4-E4B-it

google·Omni

≈5.9 GB

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.

8B128K ctx Apache 2.0
4.4MMar 2026ollama
google logo

gemma-4-12B-it

google·Omni

≈11 GB

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.

12B256K ctx Apache 2.0
1.6MMay 2026ollama
openbmb logo

MiniCPM-o-4_5

openbmb·Omni

≈7.2 GB

A Gemini 2.5 Flash Level MLLM for Vision, Speech, and Full-Duplex Mulitmodal Live Streaming on Your Phone

9.4B40K ctx Apache 2.0
155.7KFeb 2026gguf
google logo

gemma-4-E4B

google·Omni

≈5.9 GB

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.

8B128K ctx Apache 2.0
452KMar 2026gguf
google logo

gemma-4-E2B-it

google·Omni

≈3.8 GB

Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on E2B, E4B, and 12B) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants.

5.1B128K ctx Apache 2.0
2.8MMar 2026ollama
nvidia logo

music-flamingo-hf

nvidia·Omni

≈5.8 GB

Music Flamingo (MF) is a fully open, state-of-the-art Large Audio-Language Model (LALM) designed to advance music (including song) understanding in foundational audio models. MF brings together innovations in:

8.3B32K ctx Custom licence
100.7KNov 2025gguf

WeVisCut FAQ

How much VRAM does WeVisCut need?
About 7.1 GB at Q4_K_M with an 8K context. Q8_0 needs about 11 GB and full precision about 20 GB. Longer contexts need more memory for the KV cache.
Can WeVisCut run on the RTX 4060?
Yes. It fits at Q4_K_M (about 7.0 GB), at an estimated 26 tokens/s. It fits comfortably on the Apple M4 (16 GB), Radeon RX 6700 XT, RTX 3060 12GB.
Can I use WeVisCut commercially?
Unclear. It is released under Custom licence. Custom licence: read the model card before using it commercially. Check the licence on the model card before using it in a product.
What is the context length of WeVisCut?
262,144 tokens (256K). The full window needs far more memory than an 8K context.
How do I run WeVisCut with llama.cpp?
Run llama-server -hf prithivMLmods/tencent-WeVisCut-GGUF:Q4_K_M. It downloads the GGUF and starts an OpenAI-compatible server.

Data from Hugging Face, the model card and public leaderboards. Memory and speed figures are estimates; real numbers depend on the runtime, context length and drivers. Always check the licence yourself before commercial use.