NVIDIA · Blackwell · 24 GB · Qwen3.8 27B · Q4_K_M at 8K context: ~51 tok/s
NVIDIA GeForce RTX 5090 D V2 24GB
The China-market RTX 5090 with a quarter of the memory cut away: 24 GB at 1.3 TB/s on the same chip. On eBay it is mostly sold as a two-slot blower card for multi-GPU rigs. Compare the 32 GB RTX 5090 before buying: it holds a third more and often costs no more.
Lowest price right now
$5,697
eBay · free shipping · $237 per GB · observed 2026-11-14 18:37 UTC
Buy the cheapest on eBay, $5,697
free shipping · Open box · ewwi6814 77 · 97.5%
Price alert
One email per move, item plus shipping. Clear the price to hear about any drop. Manage alerts.
Price history
Daily low across every tracked vendor, last 4 days.
Where to buy
| Vendor | Price | Condition | Observed | |
|---|---|---|---|---|
| eBay cheapest active listing |
$5,697 free shipping |
new and used | 2026-11-14 18:37 UTC | Go to eBay |
On eBay now
40 listings passed our filters at the last check, 7 min ago. We check every 28 minutes. Cheapest delivered first. Seller reputation is feedback count and positive rate.
-
$5,697free shipping -
$5,700free shipping -
$5,700free shipping -
$5,700free shipping -
$5,700free shipping -
$5,850free 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+ | ~68 | ~66 | ~3,100 |
| Gemma 4 31B Mar 2026 | Q5_K_M | 24 GB | fits | 8K | ~39 | ~1,200 | |
| Gemma 4 26B-A4B Mar 2026 | Q6_K | 22 GB | fits | 87K | ~140 | ~130 | ~9,700 |
| Gemma 3 27B Mar 2025 | Q5_K_M | 22 GB | fits | 36K | ~43 | ~39 | ~1,300 |
| 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 44 GB in system RAM | ~7.7 | ~6.8 | ~1,500 | |
| gpt-oss | |||||||
| gpt-oss-120b Aug 2025 | Q4_K_M | 73 GB | runs with 50 GB in system RAM | ~17 | ~16 | ~3,600 | |
| gpt-oss-20b Aug 2025 | Q6_K | 19 GB | fits | 104K | ~140 | ~110 | ~10,000 |
| Llama | |||||||
| Llama 4 Scout 109B-A17B Apr 2025 | Q4_K_M | 70 GB | runs with 46 GB in system RAM | ~5.4 | ~4.8 | ~1,100 | |
| Llama 3.3 70B Dec 2024 | Q4_K_M | 47 GB | runs with 24 GB in system RAM | ~1.7 | ~1.3 | ~260 | |
| Llama 3.1 8B Jul 2024 | Q8_0 | 11 GB | fits | 104K | ~91 | ~68 | ~4,600 |
| 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 44 GB in system RAM | ~0.9 | ~0.8 | ~140 | |
| Mistral Small 4 119B-A6.5B Mar 2026 | Q4_K_M | 79 GB | runs with 55 GB in system RAM | ~12 | ~2,800 | ||
| Mistral Small 3.2 24B Jun 2025 | Q6_K | 23 GB | fits | 16K | ~42 | ~1,500 | |
| Nemotron | |||||||
| Nemotron 3.5 Lightning 30B-A3B Aug 2026 | Q5_K_M | 24 GB | fits | 10K | ~160 | ~11,000 | |
| Qwen | |||||||
| Qwen3.8 Flash-Next 125B-A6B Aug 2026 | Q4_K_M | 78 GB | runs with 54 GB in system RAM | ~15 | ~14 | ~3,100 | |
| Qwen3.8 27B Aug 2026 (your pick) | Q5_K_M | 21 GB | fits | 49K | ~44 | ~41 | ~1,400 |
| Qwen3.6 27B Apr 2026 | Q5_K_M | 21 GB | fits | 49K | ~44 | ~41 | ~1,400 |
| Qwen3.6 35B-A3B Apr 2026 | Q4_K_M | 23 GB | fits | 54K | ~170 | ~160 | ~12,000 |
| Qwen3.5 9B Feb 2026 | Q8_0 | 11 GB | fits | 128K+ | ~89 | ~82 | ~4,100 |
| 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 | Q4_K_M | 24 GB | fits | 9K | ~39 | ~1,100 | |
Specifications
- Memory
- 24 GB GDDR7
- Memory bandwidth
- 1,344 GB/s
- Bus width
- 384-bit
- Shader cores
- 21,760
- Tensor cores
- 680
- FP32
- 104.8 TFLOPS
- FP16 (dense)
- 210 TFLOPS
- FP8 (dense)
- 419 TFLOPS
- Power
- 575 W
- Form factor
- Two-slot blower or triple-slot cooler, by board maker
- Interface
- PCIe 5.0 x16
- Architecture
- Blackwell
- Launched
- 2025
Source: manufacturer specifications.