AMD · RDNA 3 · 20 GB · Qwen3.8 27B · Q4_K_M: ~24 tok/s
AMD Radeon RX 7900 XT 20GB
20 GB is an awkward size that happens to suit local models: room for a 24B dense model at Q6 or a 30B at Q4 with a short context, at used prices well under a 7900 XTX.
Lowest price right now
$699.99
eBay · free shipping · $34 per GB · observed 2026-10-18 02:55 UTC
Buy the cheapest on eBay, $699.99
free shipping · Used · paymore_moonvalley_arizona 1,070 · 97.1%
Price alert
One email per move, item plus shipping. Clear the price to hear about any drop. Manage alerts.
Price history
Tracking started today. The chart appears once there is a second day of data.
Where to buy
| Vendor | Price | Condition | Observed | |
|---|---|---|---|---|
| eBay cheapest active listing |
$699.99 free shipping |
new and used | 2026-10-18 02:55 UTC | Go to eBay |
On eBay now
34 listings passed our filters at the last check. Cheapest delivered first. Seller reputation is feedback count and positive rate.
-
$699.99free shipping -
$776.10$750 + $26.10 shipping -
$785.99$775 + $10.99 shipping -
$799.99free shipping -
$799.99free shipping -
$799.99free shipping
What it runs
For each model, grouped by vendor with the newest first: the largest quant that fits with 8K of context and the estimated speed for one user, at 8K and at 32K of context. On a card, a model too big for its memory can still run with part of the weights in system RAM; we assume 64 GB, and the speed shown accounts for it. The model you picked on the ladder is highlighted. Estimates from the spec sheet, not measurements: see how we estimate.
| Model | Quant | Needs at 8K | Fit | Max context | Decode at 8K | At 32K | Prefill tok/s |
|---|---|---|---|---|---|---|---|
| DeepSeek | |||||||
| DeepSeek V4.1 Flash 552B-A16B Sep 2026 | Q4_K_M | 338 GB | does not fit | ||||
| DeepSeek V4 Flash 284B-A13B Jul 2026 | Q4_K_M | 175 GB | does not fit | ||||
| DeepSeek V3.1 671B-A37B Aug 2025 | Q4_K_M | 411 GB | does not fit | ||||
| Gemma | |||||||
| Gemma 4 12B May 2026 | Q8_0 | 14 GB | fits | 128K+ | ~31 | ~30 | ~1,100 |
| Gemma 4 31B Mar 2026 | Q3_K_M | 17 GB | fits | 41K | ~25 | ~23 | ~420 |
| Gemma 4 26B-A4B Mar 2026 | Q5_K_M | 20 GB | fits | 28K | ~94 | ~3,400 | |
| Gemma 3 27B Mar 2025 | Q4_K_M | 19 GB | fits | 21K | ~23 | ~470 | |
| GLM | |||||||
| GLM-5.3 753B-A40B Aug 2026 | Q4_K_M | 462 GB | does not fit | ||||
| GLM-5.3 Flash 320B-A18B Aug 2026 | Q4_K_M | 197 GB | does not fit | ||||
| GLM-4.6 355B-A32B Sep 2025 | Q4_K_M | 221 GB | does not fit | ||||
| GLM-4.5 Air 106B-A12B Jul 2025 | Q4_K_M | 68 GB | runs with 48 GB in system RAM | ~6.9 | ~5.9 | ~540 | |
| gpt-oss | |||||||
| gpt-oss-120b Aug 2025 | Q4_K_M | 73 GB | runs with 54 GB in system RAM | ~16 | ~14 | ~1,300 | |
| gpt-oss-20b Aug 2025 | Q6_K | 19 GB | fits | 25K | ~84 | ~3,600 | |
| Llama | |||||||
| Llama 4 Scout 109B-A17B Apr 2025 | Q4_K_M | 70 GB | runs with 50 GB in system RAM | ~4.9 | ~4.3 | ~380 | |
| Llama 3.3 70B Dec 2024 | Q4_K_M | 47 GB | runs with 28 GB in system RAM | ~1.4 | ~1.1 | ~91 | |
| Llama 3.1 8B Jul 2024 | Q8_0 | 11 GB | fits | 74K | ~42 | ~31 | ~1,600 |
| MiMo | |||||||
| MiMo V2.6 Pro 1.02T-A42B Sep 2026 | Q4_K_M | 624 GB | does not fit | ||||
| MiMo V2.6 Flash 309B-A15B Sep 2026 | Q4_K_M | 190 GB | does not fit | ||||
| Mistral | |||||||
| Mistral Medium 3.5 128B Mar 2026 | Q3_K_M | 67 GB | runs with 48 GB in system RAM | ~0.8 | ~50 | ||
| Mistral Small 4 119B-A6.5B Mar 2026 | Q3_K_M | 65 GB | runs with 45 GB in system RAM | ~13 | ~990 | ||
| Mistral Small 3.2 24B Jun 2025 | Q5_K_M | 20 GB | fits | 8K | ~22 | ~540 | |
| Nemotron | |||||||
| Nemotron 3.5 Lightning 30B-A3B Aug 2026 | Q3_K_M | 17 GB | fits | 128K+ | ~130 | ~120 | ~3,700 |
| Qwen | |||||||
| Qwen3.8 Flash-Next 125B-A6B Aug 2026 | Q3_K_M | 63 GB | runs with 43 GB in system RAM | ~18 | ~17 | ~1,100 | |
| Qwen3.8 27B Aug 2026 (your pick) | Q4_K_M | 19 GB | fits | 30K | ~24 | ~480 | |
| Qwen3.6 27B Apr 2026 | Q4_K_M | 19 GB | fits | 30K | ~24 | ~480 | |
| Qwen3.6 35B-A3B Apr 2026 | Q3_K_M | 19 GB | fits | 64K | ~130 | ~110 | ~4,300 |
| Qwen3.5 9B Feb 2026 | Q8_0 | 11 GB | fits | 128K+ | ~41 | ~38 | ~1,400 |
| Qwen3 Coder 480B-A35B Jul 2025 | Q4_K_M | 296 GB | does not fit | ||||
| Qwen3 235B-A22B Jul 2025 | Q4_K_M | 146 GB | does not fit | ||||
| Qwen3 32B Apr 2025 | Q3_K_M | 20 GB | fits | 9K | ~22 | ~390 | |
Specifications
- Memory
- 20 GB GDDR6
- Memory bandwidth
- 800 GB/s
- Bus width
- 320-bit
- Shader cores
- 5,376
- FP32
- 51.6 TFLOPS
- FP16 (dense)
- 103 TFLOPS
- Power
- 315 W
- Form factor
- Dual or triple slot, full height
- Interface
- PCIe 4.0 x16
- Architecture
- RDNA 3
- Launched
- 2022
- MSRP
- $899
Source: manufacturer specifications.