GeForce RTX 4090: 24GB VRAM — What Local LLMs Can It Run?
Check prices
The GeForce RTX 4090 has 24GB of GDDR6X memory with 1,008 GB/s of bandwidth. No used-price data yet. At Q4 quantization it runs models up to 32B parameters at roughly 36 tokens/s.
Full GeForce RTX 4090 specifications
What LLMs can the GeForce RTX 4090 run?
One reference model per family (its main version)| Model | Q4_K_M | Q8_0 | FP16 | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Fits | Speed t/s | Context | Fits | Speed t/s | Context | Fits | Speed t/s | Context | |
| DeepSeekDeepSeek R1 Qwen3 8B | ✓ | 121.8 | 128K | ✓ | 71.9 | 32K | ✓ | 40.3 | 32K |
| GemmaGemma 4 31B | ✗ | — | — | ✗ | — | — | ✗ | — | — |
| GLMGLM 4.7 Flash | ✓ | 148.9 | 32K | Offload | 22.2 | — | Offload | 12.3 | — |
| gpt-ossgpt-oss 20B | ✓ | 139.4 | 128K | ✓ | 103.4 | 8K | Offload | 9.2 | — |
| KimiKimi Linear 48B A3B | Offload | 44.3 | — | Offload | 26 | — | Offload | 14.5 | — |
| LlamaLlama 3.1 8B | ✓ | 125.1 | 128K | ✓ | 73.7 | 32K | ✓ | 41.2 | 32K |
| MistralMistral 7B v0.3 | ✓ | 135.8 | 32K | ✓ | 80.6 | 32K | ✓ | 45.3 | 32K |
| QwenQwen3 8B | ✓ | 121.8 | 32K | ✓ | 71.9 | 32K | ✓ | 40.3 | 32K |
Fits when weights + KV cache + runtime overhead is at or below usable memory, single card. "Offload" on MoE models means it runs with expert weights in system RAM (only attention layers and KV cache stay in VRAM); that speed assumes 70 GB/s RAM bandwidth. The context column is the largest tier that fits.
GeForce RTX 4090 used price history
Not enough history to draw a chart yet — price tracking just started.
Prices by platform
Buy links may carry affiliate tags; they do not change the price you pay.
FAQ
How much VRAM does the GeForce RTX 4090 have?
The GeForce RTX 4090 has 24GB of GDDR6X memory with 1,008 GB/s of bandwidth.
Can the GeForce RTX 4090 run a 70B model?
Not on a single card. At Q4_K_M with an 8K context a 70B dense model needs about 43GB, more than the GeForce RTX 4090's 24GB.
How much does a used GeForce RTX 4090 cost?
No used-price data yet; it will appear here once the feed is live.