Intel Arc Pro A50 6GB
The Intel Arc Pro A50 6GB has 6 GB VRAM and 192 GB/s memory bandwidth. It can run 23 of our 99 tracked models natively in VRAM at 8k context.
With 6 GB GDDR6, the Intel Arc Pro A50 6GB is a workstation-tier GPU that can run 23 models natively.
Intel Arc Pro A50 6GB: 2022 Xe-HPG workstation GPU with 6GB ECC GDDR6 at 192 GB/s, dual-slot Arc Pro entry.
7B at Q4 natively; 6GB limits model selection. ~2-4 t/s for 7B via Vulkan.
Vulkan via llama.cpp works cross-platform. SYCL backend available with oneAPI. Primarily a professional workstation card.
| Vendor | Intel |
| Architecture | Xe-HPG (Alchemist) |
| VRAM | 6 GB |
| Memory type | GDDR6 |
| Memory bandwidth | 192 GB/s |
| Compute backend | VULKAN |
| Tier | Workstation |
| Released | 2022 |
| Models (native) | 23 / 99 |
| Models (offload) | 31 / 99 |
Software: Vulkan backend works in llama.cpp; SYCL backend available with oneAPI toolkit. Primarily a workstation/professional card.
Need more than 6GB? Rent an RTX 3080 10GB from $0.17/hr ↗
Prices as of Oct 13, 2026. Rates change; check RunPod for current pricing. Affiliate link: CanItRun may earn a commission. Doesn't affect the fit calculation above.
Popular models for this GPU
Models this GPU runs natively in VRAM (23)
- Bonsai 27B27B · MMLU-Pro ~81.51-bit (Q1_0) · ~33.9 t/s
- Gemma 4 12B (Unified)12B · MMLU-Pro 77.2Q2_K · ~30.6 t/s
- Qwen 3.5 9B9B · MMLU-Pro 82.5Q3_K_M · ~34.3 t/s
- Ornith 1.5 9B9B · MMLU-Pro N/AQ3_K_M · ~34.3 t/s
- Llama 3.1 8B Instruct8B · MMLU-Pro 48.3Q3_K_M · ~32.2 t/s
Show 18 more
- DeepSeek R1 Distill Llama 8B8B · MMLU-Pro 41.0Q3_K_M · ~32.2 t/s
- Qwen3 8B8B · MMLU-Pro 56.7Q3_K_M · ~31.2 t/s
- Qwen 2.5 7B Instruct7.6B · MMLU-Pro 56.3Q3_K_M · ~38.3 t/s
- Gemma 4 E4B7.52B · MMLU-Pro 69.4Q4_K_M · ~50.9 t/s
- Mistral 7B Instruct v0.37.25B · MMLU-Pro 30.0Q3_K_M · ~34.6 t/s
- Gemma 4 E2B4.65B · MMLU-Pro 60.0Q8_0 · ~59.2 t/s
- Gemma 3 4B Instruct4B · MMLU-Pro 43.6Q8_0 · ~33.4 t/s
- Phi-3.5 Mini Instruct3.8B · MMLU-Pro 47.4Q3_K_M · ~31.4 t/s
- Phi-4-mini Instruct3.8B · MMLU-Pro 67.3Q6_K · ~37.4 t/s
- Llama 3.2 3B Instruct3.2B · MMLU-Pro 24.0Q8_0 · ~36.5 t/s
- Qwen 2.5 3B Instruct3.1B · MMLU-Pro 32.4Q8_0 · ~42.6 t/s
- Gemma 2 2B Instruct2.6B · MMLU-Pro 17.8Q8_0 · ~43.3 t/s
- SmolLM2 1.7B Instruct1.7B · MMLU-Pro 19.0BF16 · ~32.2 t/s
- Qwen 2.5 1.5B Instruct1.5B · MMLU-Pro 16.8BF16 · ~48 t/s
- Llama 3.2 1B Instruct1.24B · MMLU-Pro 12.5BF16 · ~58 t/s
- Gemma 3 1B Instruct1B · MMLU-Pro 14.7BF16 · ~65.2 t/s
- Qwen 2.5 0.5B Instruct0.5B · MMLU-Pro 10.0BF16 · ~125.7 t/s
- SmolLM2 360M Instruct0.36B · MMLU-Pro 8.0BF16 · ~122.4 t/s
Models that fit with CPU offload (31)
These use system RAM for layers that don't fit in VRAM, so expect much slower inference.
- Mixtral 8x7B Instruct v0.146.7B · MMLU-Pro 29.7Q3_K_M · ~5.5 t/s
- Command-R 35B35B · MMLU-Pro 33.0Q3_K_M · ~1.1 t/s
- Qwen 3.5 35B-A3B (MoE)35B · MMLU-Pro 85.3Q5_K_M · ~19.5 t/s
- Qwen 3.6 35B-A3B (MoE)35B · MMLU-Pro 85.2Q5_K_M · ~19.5 t/s
- Ornith 1.5 35B-A3B (MoE)35B · MMLU-Pro N/AQ5_K_M · ~19.5 t/s
- Yi 1.5 34B Chat34.4B · MMLU-Pro 37.0Q5_K_M · ~1.1 t/s
- Qwen3 32B32.8B · MMLU-Pro 65.5Q6_K · ~1.1 t/s
- Qwen 2.5 32B Instruct32.5B · MMLU-Pro 69.0Q5_K_M · ~1.2 t/s
- Qwen 2.5 Coder 32B Instruct32.5B · MMLU-Pro 62.3Q5_K_M · ~1.2 t/s
- DeepSeek R1 Distill Qwen 32B32.5B · MMLU-Pro 65.0Q5_K_M · ~1.2 t/s
- Nemotron 3 Nano 30B32B · MMLU-Pro 78.3Q6_K · ~16.7 t/s
- Gemma 4 31B30.7B · MMLU-Pro 85.2Q6_K · ~1.1 t/s
- Qwen3 30B-A3B (MoE)30B · MMLU-Pro 61.5Q6_K · ~15.5 t/s
- Nemotron 3.5 Lightning 30B-A3B30B · MMLU-Pro 81.6Q5_K_M · ~15.4 t/s
- Muse Glimmer 30B27.8B · MMLU-Pro N/AQ6_K · ~1.4 t/s
- Gemma 2 27B Instruct27.2B · MMLU-Pro 38.0Q6_K · ~1.2 t/s
- Gemma 3 27B Instruct27B · MMLU-Pro 67.5Q6_K · ~1.3 t/s
- Qwen 3.6 27B27B · MMLU-Pro 86.2Q6_K · ~1.4 t/s
- UI-Mate 27B27B · MMLU-Pro ~86.2Q6_K · ~1.4 t/s
- Bonsai 2 27B27B · MMLU-Pro N/ATernary (Q2_0) · ~14.9 t/s
- Qwen 3.8 27B27B · MMLU-Pro N/AQ6_K · ~1.4 t/s
- Gemma 4 26B (MoE)25.2B · MMLU-Pro 82.6Q8_0 · ~10.8 t/s
- Mistral Small 3.1 24B Instruct24B · MMLU-Pro 66.8Q8_0 · ~1.1 t/s
- Mistral Small 22B22.2B · MMLU-Pro 49.2Q8_0 · ~1.2 t/s
- GPT-OSS 20B21B · MMLU-Pro 67.9Q8_0 · ~11.5 t/s
- Qwen3 14B14.8B · MMLU-Pro 61.0Q8_0 · ~2 t/s
- Qwen 2.5 14B Instruct14.7B · MMLU-Pro 63.7Q8_0 · ~2 t/s
- Phi-4 14B Instruct14B · MMLU-Pro 70.4Q8_0 · ~2.1 t/s
- Mistral Nemo 12B Instruct12.2B · MMLU-Pro 35.6BF16 · ~1.2 t/s
- Gemma 3 12B Instruct12.2B · MMLU-Pro 60.6BF16 · ~1.2 t/s
- Gemma 2 9B Instruct9.2B · MMLU-Pro 32.0BF16 · ~1.5 t/s
Too large for this GPU (45)
- Llama 3.3 70B Instruct
- Qwen 2.5 72B Instruct
- DeepSeek R1 Distill Llama 70B
- Llama 3.1 70B Instruct
- Mixtral 8x22B Instruct v0.1
- Llama 3.1 405B Instruct
- DeepSeek V3 671B
- DeepSeek R1 671B
- Llama 4 Scout 109B
- Llama 4 Maverick 400B
- Qwen3 235B-A22B (MoE)
- MiniMax M1 456B
- GPT-OSS 120B
- GLM-4.5 355B
- GLM-4.5 Air 106B
- GLM-4.6 355B
- GLM-4.6V 106B
- GLM-4.7 358B
- Qwen 3.5 122B-A10B (MoE)
- MiniMax M2.5 229B
- GLM-5 744B
- MiniMax M2.7 229B
- Nemotron 3 Super 120B
- Kimi K2.6
- GLM-5.1 754B
- DeepSeek V4 Pro 1.6T
- DeepSeek V4 Flash 284B
- Mistral Medium 3.5 128B
- GLM-5.2 753B
- Nemotron 3 Ultra 550B-A55B
- Step 3.5 Flash
- Step 3.7 Flash
- MiMo V2.5 Pro
- Kimi K2.5
- MiniMax M3
- Inkling
- Kimi K3
- DeepSeek V4 Flash 0731 284B
- Qwen3.8 2.4T-A95B
- Qwen3.8-Flash-Next
- DeepSeek V4 Pro 0813 1.6T
- Ornith 1.5 397B (MoE)
- GLM-5.3 753B
- GLM-5.3-Flash 320B
- DeepSeek V4.1 Flash 552B
Frequently asked questions
- How much VRAM does the Intel Arc Pro A50 6GB have?
- The Intel Arc Pro A50 6GB has 6 GB of GDDR6 with 192 GB/s memory bandwidth.
- What is the Intel Arc Pro A50 6GB best for?
- With 6 GB of VRAM, the Intel Arc Pro A50 6GB is best for running compact models (1B–8B) at low quantization, suitable for edge inference, prototyping, and lightweight tasks.
- What LLMs can the Intel Arc Pro A50 6GB run locally?
- The Intel Arc Pro A50 6GB can run 23 of the 99 open-weight models tracked by CanItRun natively in VRAM at 8k context. Top options include: Ornith 1.5 9B at Q3_K_M, Qwen 3.5 9B at Q3_K_M, Bonsai 27B at 1-bit (Q1_0).
- Can the Intel Arc Pro A50 6GB run Gemma 4 31B?
- The Intel Arc Pro A50 6GB can run Gemma 4 31B with CPU offload at Q6_K quantization, but inference will be slower than native VRAM execution.
- Can the Intel Arc Pro A50 6GB run Qwen 3.6 27B?
- The Intel Arc Pro A50 6GB can run Qwen 3.6 27B with CPU offload at Q6_K quantization, but inference will be slower than native VRAM execution.
- Can the Intel Arc Pro A50 6GB run Qwen3 8B?
- Yes. The Intel Arc Pro A50 6GB runs Qwen3 8B natively in VRAM at Q3_K_M quantization, achieving approximately 31.2 tokens per second.