This GPU offers reliable throughput for local AI workloads. Pair it with the right model quantization to hit your desired tokens/sec, and monitor prices below to catch the best deal.
Quick Answer: RX 6900 XT has 16GB VRAM, enough for models up to roughly 40B parameters at 4-bit quantization. It draws 300W under load.
With 16GB VRAM, RX 6900 XT can run models up to approximately 40B parameters using 4-bit quantization. That covers 13B-34B comfortably; 70B-class models only fit with heavy offloading, at much lower throughput.
Consider RTX 4090 — Double the VRAM for larger models.
Buy directly on Amazon with fast shipping and reliable customer service.
Essential accessories to pair with RX 6900 XT
Accessories Total
Typical prices — check Amazon for current pricing
💡 Not ready to buy? Try cloud GPUs first
Test RX 6900 XT performance in the cloud before investing in hardware. Pay by the hour with no commitment.
Showing 12 of 80 rows. Speeds are calculated estimates, not measurements — search for your model to jump straight to it.
| Model | Size | Quantization | Tokens/sec | VRAM used |
|---|---|---|---|---|
| Deepseek AI Deepseek Coder 1.3B Instruct | 1.3B | Q4 | ~125 tok/sEstimated | 1GB |
| Deepseek AI Deepseek R1 Distill Qwen 1.5B | 1.5B | Q4 | ~125 tok/sEstimated | 1GB |
| Deepseek AI Deepseek Ocr 2 | Unknown | Q4 | ~105 tok/sEstimated | 2GB |
| Deepseek AI Deepseek Ocr | Unknown | Q4 | ~105 tok/sEstimated | 2GB |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 8bit | 8B | Q4 | ~105 tok/sEstimated | 4GB |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 4bit | 8B | Q4 | ~105 tok/sEstimated | 4GB |
| Deepseek AI Deepseek R1 Distill Qwen 7B | 7B | Q4 | ~105 tok/sEstimated | 4GB |
| Nineninesix Kani Tts 2 En | Unknown | Q4 | ~99 tok/sEstimated | 1GB |
| Qwen Qwen3 Tts 12hz 1 7B Customvoice | 7B | Q4 | ~99 tok/sEstimated | 1GB |
| Zai Org Glm Ocr | Unknown | Q4 | ~99 tok/sEstimated | 1GB |
| Qwen Qwen3 Asr 1 7B | 7B | Q4 | ~99 tok/sEstimated | 2GB |
| Nari Labs Dia2 2B | 2B | Q4 | ~99 tok/sEstimated | 1GB |
Showing 12 of 240 rows.
| Model | Size | Quantization | Verdict | Estimated speed | VRAM needed |
|---|---|---|---|---|---|
| 01 AI Yi 1 5 34B Chat | 34B | Q4 | Not supported | ~29 tok/sEstimated | 18GB (have 16GB) |
| 01 AI Yi 1 5 34B Chat | 34B | Q8 | Not supported | ~20 tok/sEstimated | 35GB (have 16GB) |
| 01 AI Yi 1 5 34B Chat | 34B | FP16 | Not supported | ~11 tok/sEstimated | 69GB (have 16GB) |
| AI Forever Rugpt 3.5 13B | 13B | Q4 | Fits comfortably | ~62 tok/sEstimated | 7GB (have 16GB) |
| AI Forever Rugpt 3.5 13B | 13B | Q8 | Fits comfortably | ~43 tok/sEstimated | 13GB (have 16GB) |
| AI Forever Rugpt 3.5 13B | 13B | FP16 | Not supported | ~24 tok/sEstimated | 26GB (have 16GB) |
| AI Mo Kimina Prover 72B | 72B | Q4 | Not supported | ~17 tok/sEstimated | 37GB (have 16GB) |
| AI Mo Kimina Prover 72B | 72B | Q8 | Not supported | ~12 tok/sEstimated | 73GB (have 16GB) |
| AI Mo Kimina Prover 72B | 72B | FP16 | Not supported | ~6.3 tok/sEstimated | 146GB (have 16GB) |
| Alibaba Nlp Gte Qwen2 1.5B Instruct | 1.5B | Q4 | Fits comfortably | ~99 tok/sEstimated | 1GB (have 16GB) |
| Alibaba Nlp Gte Qwen2 1.5B Instruct | 1.5B | Q8 | Fits comfortably | ~69 tok/sEstimated | 2GB (have 16GB) |
| Alibaba Nlp Gte Qwen2 1.5B Instruct | 1.5B | FP16 | Fits comfortably | ~38 tok/sEstimated | 4GB (have 16GB) |
Note: Performance estimates are calculated. Real results may vary. Methodology · Submit real data
Explore how RTX 4090 stacks up for local inference workloads.
Explore how RTX 4080 stacks up for local inference workloads.
Explore how RTX 4070 Ti stacks up for local inference workloads.
Explore how RTX 3090 stacks up for local inference workloads.
Explore how RX 7900 XTX stacks up for local inference workloads.
RPG • 2020
RPG • 2023
Action RPG • 2023
RPG • 2023
Survival Horror • 2023
Action RPG • 2022
Action RPG • 2024
Action Adventure • 2025
Survival Horror • 2023
Action • 2022
Action Adventure • 2023
Action Adventure • 2019