NVIDIA GeForce RTX 4090
NVIDIA · 24GBGDDR6X · Can run 84 models
BuyAmazon
| Manufacturer | NVIDIA |
| VRAM | 24 GB |
| Memory Type | GDDR6X |
| Architecture | Ada Lovelace |
| CUDA Cores | 16,384 |
| Tensor Cores | 512 |
| Bandwidth | 1008 GB/s |
| TDP | 450W |
| MSRP | $1,599 |
| Released | Oct 12, 2022 |
AI Notes
The RTX 4090 remains one of the best GPUs for local AI inference. Its 24GB of GDDR6X VRAM can run 13B models at full precision and 30B+ models with quantization. The massive tensor core count delivers class-leading inference throughput among consumer GPUs.
Compatible Models
| Model | Parameters | Best Quant | VRAM Used | Fit | Est. Speed |
|---|---|---|---|---|---|
| Qwen 3 0.6B | 600M | Q4_K_M | 2.5 GB | Runs | ~403 tok/s |
| Qwen 3.5 0.8B | 800M | Q4_K_M | 1.5 GB | Runs | ~672 tok/s |
| Gemma 3 1B | 1B | Q8_0 | 2 GB | Runs | ~504 tok/s |
| Llama 3.2 1B | 1B | Q8_0 | 3 GB | Runs | ~336 tok/s |
| DeepSeek R1 1.5B | 1.5B | Q8_0 | 3 GB | Runs | ~336 tok/s |
| SmolLM2 1.7B | 1.7B | Q8_0 | 2.7 GB | Runs | ~373 tok/s |
| Gemma 2 2B | 2B | Q8_0 | 4 GB | Runs | ~252 tok/s |
| Gemma 3n E2B | 2B | Q4_K_M | 3.3 GB | Runs | ~305 tok/s |
| Gemma 4 E2B | 2B | Q4_K_M | 4 GB | Runs | ~252 tok/s |
| Qwen 3.5 2B | 2B | Q4_K_M | 3 GB | Runs | ~336 tok/s |
| Llama 3.2 3B | 3B | Q8_0 | 5 GB | Runs | ~202 tok/s |
| StarCoder2 3B | 3B | Q4_K_M | 3.5 GB | Runs | ~288 tok/s |
| Phi-3 Mini 3.8B | 3.8B | Q8_0 | 5.8 GB | Runs | ~174 tok/s |
| Phi-4 Mini 3.8B | 3.8B | Q4_K_M | 4.5 GB | Runs | ~224 tok/s |
| Gemma 3 4B | 4B | Q4_K_M | 5 GB | Runs | ~202 tok/s |
| Gemma 3n E4B | 4B | Q4_K_M | 4.5 GB | Runs | ~224 tok/s |
| Gemma 4 E4B | 4B | Q4_K_M | 6 GB | Runs | ~168 tok/s |
| Qwen 3 4B | 4B | Q4_K_M | 4.5 GB | Runs | ~224 tok/s |
| Qwen 3.5 4B | 4B | Q4_K_M | 4.5 GB | Runs | ~224 tok/s |
| Yi 1.5 6B | 6B | Q4_K_M | 5 GB | Runs | ~202 tok/s |
| Codestral Mamba 7B | 7B | Q4_K_M | 6.9 GB | Runs | ~146 tok/s |
| DeepSeek R1 7B | 7B | Q8_0 | 9 GB | Runs | ~112 tok/s |
| Falcon 3 7B | 7B | Q4_K_M | 6.8 GB | Runs | ~148 tok/s |
| InternLM 2.5 7B | 7B | Q4_K_M | 5.5 GB | Runs | ~183 tok/s |
| Mistral 7B | 7B | Q8_0 | 9 GB | Runs | ~112 tok/s |
| OpenChat 3.5 7B | 7B | Q4_K_M | 6.9 GB | Runs | ~146 tok/s |
| Qwen 2.5 7B | 7B | Q8_0 | 9 GB | Runs | ~112 tok/s |
| Qwen 2.5 Coder 7B | 7B | Q8_0 | 9 GB | Runs | ~112 tok/s |
| Qwen 2.5 VL 7B | 7B | Q4_K_M | 7 GB | Runs | ~144 tok/s |
| StarCoder2 7B | 7B | Q4_K_M | 5.5 GB | Runs | ~183 tok/s |
| WizardLM 2 7B | 7B | Q4_K_M | 6.9 GB | Runs | ~146 tok/s |
| Aya Expanse 8B | 8B | Q4_K_M | 6.5 GB | Runs | ~155 tok/s |
| Cogito 8B | 8B | Q4_K_M | 7.5 GB | Runs | ~134 tok/s |
| DeepSeek R1 8B | 8B | Q4_K_M | 7.5 GB | Runs | ~134 tok/s |
| Dolphin 3 8B | 8B | Q4_K_M | 6 GB | Runs | ~168 tok/s |
| Granite 3.3 8B | 8B | Q8_0 | 10 GB | Runs | ~101 tok/s |
| Llama 3.1 8B | 8B | Q8_0 | 10 GB | Runs | ~101 tok/s |
| Nemotron 3 Nano 8B | 8B | Q4_K_M | 7.5 GB | Runs | ~134 tok/s |
| Nous Hermes 2 8B | 8B | Q4_K_M | 6 GB | Runs | ~168 tok/s |
| Qwen 3 8B | 8B | Q4_K_M | 7.5 GB | Runs | ~134 tok/s |
| Gemma 2 9B | 9B | Q8_0 | 11 GB | Runs | ~92 tok/s |
| Qwen 3.5 9B | 9B | Q4_K_M | 7.5 GB | Runs | ~134 tok/s |
| Yi 1.5 9B | 9B | Q4_K_M | 6.5 GB | Runs | ~155 tok/s |
| Yi Coder 9B | 9B | Q4_K_M | 8 GB | Runs | ~126 tok/s |
| Falcon 3 10B | 10B | Q4_K_M | 8.5 GB | Runs | ~119 tok/s |
| Llama 3.2 Vision 11B | 11B | Q4_K_M | 8.5 GB | Runs | ~119 tok/s |
| Gemma 3 12B | 12B | Q4_K_M | 10.5 GB | Runs | ~96 tok/s |
| Mistral Nemo 12B | 12B | Q4_K_M | 9.5 GB | Runs | ~106 tok/s |
| DeepSeek R1 14B | 14B | Q4_K_M | 9.9 GB | Runs | ~102 tok/s |
| Phi-4 14B | 14B | Q4_K_M | 9.9 GB | Runs | ~102 tok/s |
| Phi-4 Reasoning 14B | 14B | Q4_K_M | 11 GB | Runs | ~92 tok/s |
| Qwen 2.5 14B | 14B | Q4_K_M | 9.9 GB | Runs | ~102 tok/s |
| Qwen 2.5 Coder 14B | 14B | Q4_K_M | 12 GB | Runs | ~84 tok/s |
| Qwen 3 14B | 14B | Q4_K_M | 12 GB | Runs | ~84 tok/s |
| StarCoder2 15B | 15B | Q8_0 | 17 GB | Runs | ~59 tok/s |
| InternLM 2.5 20B | 20B | Q4_K_M | 12 GB | Runs | ~84 tok/s |
| gpt-oss 20B | 21B | MXFP4 | 15 GB | Runs | ~67 tok/s |
| Codestral 22B | 22B | Q4_K_M | 14.7 GB | Runs | ~69 tok/s |
| Devstral 24B | 24B | Q4_K_M | 17 GB | Runs | ~59 tok/s |
| Magistral Small 24B | 24B | Q4_K_M | 17 GB | Runs | ~59 tok/s |
| Mistral Small 3.1 24B | 24B | Q4_K_M | 18 GB | Runs | ~56 tok/s |
| Gemma 4 26B | 26B | Q4_K_M | 20 GB | Runs | ~50 tok/s |
| Gemma 2 27B | 27B | Q4_K_M | 17.7 GB | Runs | ~57 tok/s |
| Gemma 3 27B | 27B | Q4_K_M | 20 GB | Runs | ~50 tok/s |
| Qwen 3.5 27B | 27B | Q4_K_M | 19 GB | Runs | ~53 tok/s |
| Qwen 3.6 27B | 27B | Q4_K_M | 20 GB | Runs | ~50 tok/s |
| Nous Hermes 2 34B | 34B | Q4_K_M | 19 GB | Runs | ~53 tok/s |
| Qwen 3.5 35B A3B | 35B | Q4_K_M | 12 GB | Runs | ~84 tok/s |
| Qwen 3 30B-A3B (MoE) | 30B | Q4_K_M | 22 GB | Runs (tight) | ~46 tok/s |
| Gemma 4 31B | 31B | Q4_K_M | 22 GB | Runs (tight) | ~46 tok/s |
| Aya Expanse 32B | 32B | Q4_K_M | 22 GB | Runs (tight) | ~46 tok/s |
| Cogito 32B | 32B | Q4_K_M | 21.5 GB | Runs (tight) | ~47 tok/s |
| DeepSeek R1 32B | 32B | Q4_K_M | 20.7 GB | Runs (tight) | ~49 tok/s |
| Qwen 2.5 32B | 32B | Q4_K_M | 20.7 GB | Runs (tight) | ~49 tok/s |
| QwQ 32B | 32B | Q4_K_M | 21.5 GB | Runs (tight) | ~47 tok/s |
| Laguna XS 2.1 | 33B | Q4_K_M | 22 GB | Runs (tight) | ~46 tok/s |
| WizardCoder 33B | 33B | Q4_K_M | 22 GB | Runs (tight) | ~46 tok/s |
| Yi 1.5 34B | 34B | Q4_K_M | 21 GB | Runs (tight) | ~48 tok/s |
| Command R 35B | 35B | Q4_K_M | 22.5 GB | Runs (tight) | ~45 tok/s |
| Qwen 2.5 Coder 32B | 32B | Q4_K_M | 23 GB | CPU Offload | ~13 tok/s |
| Qwen 3 32B | 32B | Q4_K_M | 23 GB | CPU Offload | ~13 tok/s |
| Qwen 3.6 35B-A3B | 35B | Q4_K_M | 27 GB | CPU Offload | ~11 tok/s |
| Dolphin Mixtral 8x7B | 47B | Q4_K_M | 26 GB | CPU Offload | ~12 tok/s |
| Mixtral 8x7B | 47B | Q4_K_M | 29.7 GB | CPU Offload | ~10 tok/s |
30 model(s) are too large for this hardware.