Youtu Parsing Omni — Hardware Requirements & GPU Compatibility
Vision
Youtu-Parsing-Omni is Tencent's compact omni-modal parsing model with about 5.3 billion parameters. Given a document page, natural image, chart, flowchart, geometry figure, audio clip or audio-visual video, it produces a single structured JSON output covering layout elements, text, tables, formulas, bounding boxes, timestamps, speech recognition and captions, with the output type chosen by the task prompt. It ships custom code and a vLLM plugin, so it needs the project's own tooling to run. At this size it fits on a single consumer GPU when quantized and on a mid-range card in BF16. The model configuration lists a context length of 1,048,576 tokens. It is released under Tencent's own Youtu-Parsing license, which has its own terms rather than a standard open-source license. Published in October 2026, it follows Tencent's earlier Youtu-VL and Youtu-LLM work, which the card acknowledges.
50 downloads
48 likesSpecifications
- Publisher
- Tencent
- Parameters
- 5.3B
- Architecture
- YoutuVITAForCausalLM
- Context Length
- 1,048,576 tokens
- Vocabulary Size
- 133,632
- Release Date
- 2026-10-09
- License
- Other
Get Started
HuggingFace
Run in cloud
Fits on RTX A4000 (16 GB) (4 GB headroom) · BF16
- Generation speed
- ~26 tok/s
- generation speed
- Cost per 1M output tokens
- $1.02
- per 1M output tokens
How Much VRAM Does Youtu Parsing Omni Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16est. | 16.00 | 11.2 GB | 118.3 GB | 10.67 GB | Brain floating point 16 — preferred for training |
est.= calculated VRAM estimate; no published GGUF file found for that quantization yet. Other rows are verified against real community uploads.
Which GPUs Can Run Youtu Parsing Omni?
BF16 · 11.2 GBYoutu Parsing Omni (BF16) requires 11.2 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 15+ GB is recommended. Using the full 1049K context window can add up to 107.2 GB, bringing total usage to 118.3 GB. 36 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 5080.
Runs great
— Plenty of headroom
Decent
— Enough VRAM, may be tightWhich Devices Can Run Youtu Parsing Omni?
BF16 · 11.2 GB48 devices with unified memory can run Youtu Parsing Omni, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Mini M4 (16 GB).
Runs great
— Plenty of headroomFrequently Asked Questions
- How much VRAM does Youtu Parsing Omni need?
Youtu Parsing Omni requires 11.2 GB of VRAM at BF16. Full 1049K context adds up to 107.2 GB (118.3 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 5.3B × 16 bits ÷ 8 = 10.7 GB
KV Cache + Overhead ≈ 0.5 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 107.6 GB (at full 1049K context)
VRAM usage by quantization
BF1611.2 GBBF16 + full context118.3 GB- Can I run Youtu Parsing Omni on a Mac?
Yes — Mac Mini M4 (16 GB) and 28 other Macs can run Youtu Parsing Omni. Apple Silicon uses unified memory, so the model shares RAM with the system. At BF16 you need at least 11.2 GB of usable unified memory (RAM minus macOS overhead).
- Can I run Youtu Parsing Omni locally?
Yes — Youtu Parsing Omni can run locally on consumer hardware. At BF16 quantization it needs 11.2 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Youtu Parsing Omni?
At BF16, Youtu Parsing Omni can reach ~429 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~59 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 11.18 × 0.65 = ~465 tok/s
Estimated speed at BF16 (11.2 GB)
~465 tok/s~59 tok/s~465 tok/s~429 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Youtu Parsing Omni?
At BF16, the download is about 10.67 GB.
- Which GPUs can run Youtu Parsing Omni?
36 consumer GPUs can run Youtu Parsing Omni at BF16 (11.2 GB). Top options include AMD Radeon RX 7900 XT, AMD Radeon RX 7900 XTX, NVIDIA GeForce RTX 3090, AMD Radeon RX 6700 XT. 8 GPUs have plenty of headroom for comfortable inference.
- Which devices can run Youtu Parsing Omni?
48 devices with unified memory can run Youtu Parsing Omni at BF16 (11.2 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Asus ROG Flow Z13 (2025, Ryzen AI Max+ 395, 128 GB), Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.