AMD Radeon AI PRO R9700 20 GB
Live eBay prices, cheapest first. The best one right now is $1,900 (used).
Other sizes: 32 GB
Specs
- VRAM
- 20 GB GDDR6
- Bandwidth
- 640 GB/s
- FP32
- 47.8 TFLOPS
- FP16
- 191 TFLOPS
- Power
- 300 W
- Connector
- PCIe 5.0
- Released
- 2025
- Launch price
- $1,299
- Llama 3.1 8B at Q4 (est.)
- 46–79 tokens/s
Price right now
Live on eBayFrom the live listings below; shipping included.
Listings
Live on eBay| Listing | Condition | Price | Checked |
|---|---|---|---|
| Used | $1,900or best offershipping not listed |
Auctions ending soon
No auctions ending soon right now.
Prices include shipping and were read from eBay at the time shown. They can change: check the listing before you buy.
Owner reviews
No owner reviews yet. Own one? Tell others how it runs.
From signed-in owners, checked before they're published. PerGB doesn't pay for or edit reviews.
How it compares
Live on eBay- Software: ROCm (on AMD's support list).
As much VRAM, about as fast, for less
- AMD Radeon RX 7900 XTX: $900.00, 133% of the GeForce RTX 3090, 24 GB
- NVIDIA Tesla V100 PCIe 32 GB: $975.00, 123% of the GeForce RTX 3090, 32 GB
- NVIDIA Tesla V100S PCIe: $1,099, 149% of the GeForce RTX 3090, 32 GB
Faster, for the least
- AMD Instinct MI100: $1,240, 185% of the GeForce RTX 3090, 32 GB
- NVIDIA GeForce RTX 4090: $2,094, 158% of the GeForce RTX 3090, 24 GB
- AMD Instinct MI210: $3,799, 211% of the GeForce RTX 3090, 64 GB
More VRAM, for not much more
- NVIDIA Tesla P40: $270.00, 35% of the GeForce RTX 3090, 24 GB
- AMD Instinct MI50 32 GB: $633.68, 64% of the GeForce RTX 3090, 32 GB
- NVIDIA GeForce RTX 2080 Ti 22 GB: $665.00, 71% of the GeForce RTX 3090, 22 GB
From the live listings of every GPU, shipping included; performance against the GeForce RTX 3090 (FP16 or FP32 compute and memory bandwidth). Updated with every build.
Local AI: what it runs
| Model (Q4, 8K context) | Radeon AI PRO R9700 20 GB needed | Costs now | Speed (est.) |
|---|---|---|---|
| Llama 3.1 8B | 1 card | $1,900 | 46–79 tokens/s |
| Qwen3 14B | 1 card | $1,900 | 25–43 tokens/s |
| gpt-oss 20B | 1 card | $1,900 | 62–106 tokens/s |
| Gemma 3 27B | 2 cards | not enough listings | 13–23 tokens/s |
| Qwen3 32B | 2 cards | not enough listings | 11–19 tokens/s |
| Llama 3.3 70B | 3 cards | not enough listings | 5.2–9.0 tokens/s |
| gpt-oss 120B | 4 cards | not enough listings | 43–75 tokens/s |
| Qwen3 235B-A22B | 8 cards | not enough listings | 10–17 tokens/s |
| DeepSeek R1 / V3 671B | more than 8 | — | — |
Cost: that many of the cheapest live listings. Speed: an estimate from the memory bandwidth for one person chatting. Try other models, sizes and contexts.
Decision models
| Model (Q4, 8K context) | Radeon AI PRO R9700 20 GB needed | Costs now | Per question (est.) |
|---|---|---|---|
| Julia-10.144B | 1 card | $1,900 | 7.5–16 ms |
| Laya0.421B | 1 card | $1,900 | 12–27 ms |
| Kev-4B4B | 1 card | $1,900 | 30–64 ms |
| lev4B | 1 card | $1,900 | 90–192 ms |
| OpenJev27B | 1 card | $1,900 | 107–230 ms |
| Clef27B | 1 card | $1,900 | 107–230 ms |
| Perplexity Decider v1.227B | 1 card | $1,900 | 107–230 ms |
| mammoth-127B | 1 card | $1,900 | 107–230 ms |
Decision models score the options you give them instead of writing text, so they're timed per question. The time is scaled from each model's reported speed on an RTX PRO 6000 by this card's compute and memory bandwidth, within about half either way. About the models.
Compare elsewhere
Search other shops for AMD Radeon AI PRO R9700 20 GB. These open the shop's own search; we don't show their prices here.
