Hy4 Guide: VRAM Requirements, GGUF Builds
Everything Local Model Watch has published about the Hy4 family: 2 article(s) covering the base model and its fine-tunes, plus converted builds we tracked after publication. Memory requirements below are computed by this site from file sizes, not quoted from model cards. Part of our model family index.
At a Glance
| Item | Value |
|---|---|
| Base model(s) | tencent/Hy4-base, tencent/Hy4-flash |
| Publisher | Tencent |
| License (model card) | tencent-hy-license |
| Articles | 2 |
Hardware Requirements
Estimated requirements (calculated by Local Model Watch) — 771.1B parameters
| Your VRAM | Quantization | File size | Est. memory needed |
|---|---|---|---|
| More than 1724GB of VRAM (multi-GPU or CPU offload required) | BF16 | 1436.4GB | 1723.7GB |
Memory estimates add a 20% runtime overhead (KV cache, etc.) to the actual size of the distributed files. Actual usage varies with context length, batch size and inference engine. These figures are computed by this site from file sizes, not published by the model’s authors. Compare with other models in our VRAM quick reference. What the quantization names mean: glossary.
Figures are for tencent/Hy4-base. Other sizes in this family are covered in the articles below.
Can You Run It Locally?
Not usable in Ollama, LM Studio and llama.cpp yet.
The publisher ships safetensors only, and llama.cpp’s registry does not list this architecture. llama.cpp would need to add support before these tools can run it. Today it can be run with transformers or vLLM, using the memory figures in the table above.
License — tencent-hy-license: A custom license from the publisher. Check the original terms directly, including whether commercial use is permitted.
Compiled by this site’s code from the published formats, converted builds we have found, and each engine’s own model registry. “Not found" means we have not seen such a build, not that none exists. License summaries are not legal advice — check the publisher’s original terms before relying on them.
This assessment is for tencent/Hy4-base.
Quantized and Converted Variants
→ Scroll horizontally to see all columns
| Added | Publisher | Format | Repository | Smallest VRAM tier (build, est. memory) |
|---|---|---|---|---|
| 2026-11-19 | autotrust | GGUF | autotrust/hy4-flash-smart | NVFP4 139.3GB (does not fit a single consumer GPU) |
| 2026-11-19 | autotrust | GGUF | autotrust/hy4-flash-concise | NVFP4 139.3GB (does not fit a single consumer GPU) |
| 2026-11-18 | AtomicChat | GGUF (imatrix) | AtomicChat/Hy4-flash-GGUF | IQ2_XXS 94.7GB (does not fit a single consumer GPU) |
File sizes of each build:
- Available builds in autotrust/hy4-flash-smart: NVFP4 116.1GB
- Available builds in autotrust/hy4-flash-concise: NVFP4 116.1GB
- Available builds in AtomicChat/Hy4-flash-GGUF: IQ2_XXS 78.9GB
This section is appended automatically by Local Model Watch when a converted build of this model appears after publication. Memory figures are estimated from the size of the distributed files.
Articles (the family’s own models first, then newest)
| Published | Model | Type | Article |
|---|---|---|---|
| 2026-11-20 | tencent/Hy4-base | New Models | Hy4 Model Specifications, Benchmarks, and Use Cases |
| 2026-11-18 | tencent/Hy4-flash-base | New Models | Tencent Releases Hy4 flash: A Cost-Effective Multimodal MoE Model |
Repositories
Last updated 2026-11-20 (JST). Assembled by code from our article log; no text on this page is written by an AI model.