Mixtral 8x22B needs ~99.9 GB VRAM. RTX PRO 6000 Blackwell Server Edition 96GB has 96.0 GB. With Q4_K_M quantization, expect ~23 tok/s.
Operating mode
Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.
Current mode
Balanced
Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.
Select quantization to explore
3.9 GB over capacity — needs offload or smaller quantization
Fit status
Runs with offload (needs ~3.4 GB host RAM)
Decode
22.8 tok/s
TTFT
8476 ms
Safe context
4K
Memory
99.9 GB / 96.0 GB
This setup is broadly balanced for this model.
Very little memory headroom
You can run the model, but there is not much room left for longer context, bigger batches, extra apps, or future model updates.
Buy headroom, not only minimum fit
A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | B | Runs with offload (needs ~1.9 GB host RAM) | 23.7 tok/s | 4464 ms | 4K |
| Coding | B | Runs with offload (needs ~3.4 GB host RAM) | 22.8 tok/s | 8476 ms | 4K |
| Agentic Coding | C | Runs with offload (needs ~6.1 GB host RAM) | 21.3 tok/s | 13205 ms | 4K |
| Reasoning | B | Runs with offload (needs ~3.4 GB host RAM) | 22.8 tok/s | 10018 ms | 4K |
| RAG | C | Runs with offload (needs ~6.1 GB host RAM) | 21.3 tok/s | 16506 ms | 4K |
How Mixtral 8x22B (141B params) fits at each quantization level on RTX PRO 6000 Blackwell Server Edition 96GB (96.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 55.0 GB | Low | B61 |
Q3_K_SBest for your GPU | 3 | 69.1 GB | Low | B61 |
NVFP4 | 4 | 79.0 GB | Medium | F0 |
Q4_K_M | 4 | 86.0 GB | Medium | F0 |
Q5_K_M | 5 | 101.5 GB | High | F0 |
Q6_K | 6 | 115.6 GB | High | F0 |
Q8_0 | 8 | 150.9 GB | Very High | F0 |
F16 | 16 | 289.0 GB | Maximum | F0 |
Copy-paste commands to run Mixtral 8x22B on your machine.
Run
ollama run mixtral:8x22bOpções de upgrade
NVIDIA H200 141GBOpção econômicaRaises estimated decode speed by about 327%.
Adds memory headroom for longer context windows and future model growth.
~$30,000 MSRP
NVIDIA H200 PCIe 141GBMelhor custo-benefícioRaises estimated decode speed by about 327%.
Adds memory headroom for longer context windows and future model growth.
~$30,000 MSRP
NVIDIA B200 180GBUpgrade NVIDIARaises estimated decode speed by about 612%.
Adds memory headroom for longer context windows and future model growth.
~$30,000 MSRP
Yes, RTX PRO 6000 Blackwell Server Edition 96GB can run Mixtral 8x22B with a B grade (Runs with offload (needs ~3.4 GB host RAM)). Expected decode speed: 22.8 tok/s.
Mixtral 8x22B (141B parameters) requires approximately 99.9 GB of memory with Q4_K_M quantization.
The recommended quantization for Mixtral 8x22B is Q4_K_M, which balances quality and memory efficiency.
On RTX PRO 6000 Blackwell Server Edition 96GB, Mixtral 8x22B achieves approximately 22.8 tokens per second decode speed with a time-to-first-token of 8476ms using Q4_K_M quantization.
For coding workloads, Mixtral 8x22B on RTX PRO 6000 Blackwell Server Edition 96GB receives a B grade with 22.8 tok/s and 4K context.
On RTX PRO 6000 Blackwell Server Edition 96GB, Mixtral 8x22B can safely use up to 4K tokens of context. The model's official context limit is 66K, but available memory constrains the safe maximum.
Buy headroom, not only minimum fit. A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/mixtral-8x22b-on-rtx-pro-6000-blackwell-server-96gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview: