Can the NVIDIA RTX 5090 run Mistral 7B v0.3?
Yes
QuantFP16Full quality
Memory16.2 GBof 32 GB
Speed~74tokens/s
Yes. The NVIDIA RTX 5090 can run Mistral 7B v0.3 fully in VRAM at FP16 (full quality), using about 16.2 GB of the 32 GB available, at roughly 74 tokens/second.
Try other context lengths in the calculator →Memory needed by quantization
At 8,192 tokens of context.
| Quant | Quality | Memory | Fits? | Speed |
|---|---|---|---|---|
| FP16 | Full quality | 16.2 GB | Fits | ~74 tok/s |
| Q8_0 | Near-lossless | 9.4 GB | Fits | ~140 tok/s |
| Q6_K | Excellent | 7.6 GB | Fits | ~181 tok/s |
| Q5_K_M | Very good | 6.8 GB | Fits | ~209 tok/s |
| Q4_K_M | Good, most popular | 6.1 GB | Fits | ~245 tok/s |
| Q3_K_M | Noticeably degraded | 5.2 GB | Fits | ~303 tok/s |
| Q2_K | Heavily degraded | 4.4 GB | Fits | ~395 tok/s |
How to run Mistral 7B v0.3 on the NVIDIA RTX 5090
With Ollama installed, one command downloads and starts it:
ollama run mistral:7bOllama's default download is Q4_K_M (about 6.1 GB here). You have room for FP16, which is full quality: pick that GGUF in LM Studio or from Hugging Face.
GPUs that run Mistral 7B v0.3 comfortably
- NVIDIA RTX 4060 (8 GB)
- NVIDIA RTX 3070 (8 GB)
- NVIDIA RTX 3080 10GB
- NVIDIA RTX 3060 12GB
- NVIDIA RTX 4070 (12 GB)
Models that run well on the NVIDIA RTX 5090
Run it in the cloud
Don't want to buy hardware? Run Mistral 7B v0.3 on a rented GPU today.
Rent a cloud GPU by the hour. No upfront cost, cancel any time.