How much VRAM does Intel UHD 730 have?

Intel UHD 730 ships with 0GB of VRAM. Of that, roughly 0GB is available for model weights after driver and OS overhead.

Can Intel UHD 730 run Llama 3.1 70B Instruct?

Llama 3.1 70B Instruct requires more VRAM than Intel UHD 730 provides at any quantization. Consider a smaller model, a multi-GPU setup, or a cloud GPU.

intelintegrated

Intel UHD 730 AI Model Compatibility

Name: Intel UHD 730
Brand: INTEL

What AI models can you run on a Intel UHD 730? With 0GB VRAM, this card runs 0 of 109 models in our database. Below: full grades, recommended quantizations, and tokens-per-second estimates for every model.

VRAM

0GB

Excellent fit

S + A grade

Will run

B + C grade

Too large

109

Cannot run any quant

Language Models0 of 47 run

Grade

Model

Best quant · VRAM

Speed

Experience

Llama 3.1 70B Instruct

DeepSeek R1 Distill 8B

Llama 3.1 8B Instruct

7.7B · Shanghai AI Lab

Mistral 7B Instruct v0.3

Llama 3.2 3B Instruct

DeepSeek R1 Distill 1.5B

Llama 3.2 1B Instruct

Code Models0 of 16 run

Grade

Model

Best quant · VRAM

Speed

Experience

Code Llama 13B Instruct

Multimodal & Vision0 of 6 run

Grade

Model

Best quant · VRAM

Speed

Experience

Image Generation0 of 9 run

Grade

Model

Best quant · VRAM

Speed

Experience

FLUX.1 Schnell (GGUF)

12B · Black Forest Labs

12B · Black Forest Labs

Q5_0 · 14GB

—

Cannot run

Stable Diffusion XL (CoreML)

Stable Diffusion 3 Medium (GGUF)

Stable Diffusion 2.1 Base (CoreML)

0.86B · Stability AI / Apple

CoreML-Palettized · 1.56GB

—

Cannot run

Stable Diffusion 1.5 (CoreML)

0.86B · Runway

CoreML-Palettized · 2.5GB

—

Cannot run

Stable Diffusion 1.5 (GGUF)

0.86B · Runway / GPUStack

Q4_0 · 2.13GB

—

Cannot run

Stable Diffusion 2.1 (GGUF)

Speech Recognition0 of 9 run

Grade

Model

Best quant · VRAM

Speed

Experience

Whisper Large v3 Turbo

Distil-Whisper Large v3

Whisper Tiny English (Quantized)

Text-to-Speech0 of 14 run

Grade

Model

Best quant · VRAM

Speed

Experience

Piper TTS - Amy (English)

Piper TTS - Lessac (English)

Piper TTS - LibriTTS-R (English)

Piper TTS - Spanish (MLS)

Piper TTS - French (Siwis)

Piper TTS - German (Thorsten)

Piper TTS - Chinese (Huayan)

Piper TTS - Japanese (Kokoro)

Piper TTS - Russian (Irina)

Piper TTS - Portuguese (Faber)

Piper TTS - Italian (Riccardo)

Piper TTS - Arabic (Kareem)

Audio Generation0 of 1 run

Grade

Model

Best quant · VRAM

Speed

Experience

Embedding Models0 of 5 run

Grade

Model

Best quant · VRAM

Speed

Experience

Nomic Embed Text v1.5

Snowflake Arctic Embed S

0.023B · Sentence Transformers

Q8_0 · 0.1GB

—

Cannot run

Reranker Models0 of 2 run

Grade

Model

Best quant · VRAM

Speed

Experience

Jina Reranker Tiny EN

How these grades work

Grades are computed from the ratio of Intel UHD 730's effective VRAM (0GB) to each model's required VRAM at its highest-quality quantization that still fits. S: comfortable headroom (1.5×+). A: smooth (1.2×+). B: tight but works (1.0×+). C: partial offload (0.8×+). D: heavy offload (0.5×+). F: cannot run.

Tokens-per-second figures are based on real community benchmarks (llama.cpp discussions, MLX, vLLM) scaled to model size. Real-world numbers vary with batch size, context length, and driver version.