Mistral Medium 3.5 128B — Hardware Requirements & GPU Compatibility
Chat
Mistral Medium 3.5 128B is a 127.7B-parameter open language model from Mistral AI in the Mistral family. It supports a context window of up to 262,144 tokens. At Q4_K_M it needs about 77.66 GB of VRAM — see which GPUs and Macs can run it below.
Mistral Medium 3.5 128B runs on a 96 GB+ Mac or a workstation GPU
Best local pick → Mac Studio (M3 Ultra, 96GB) · Q4_K_M · 82.8 GB · ~7 tok/s
Macs from 64 GB (Q2_K · ~10 tok/s)
Other
Not set up for it?
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Mistral Medium 3.5 128B
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
ollama run mistral-medium-3.5:128bQ4_K_M needs offload on the RTX 4090 · 58.8 GB short · 16K context
Download: Weights on Hugging Face
Other ways
each opens step-by-stepRun in cloud
Fits on H200 (141 GB) (58 GB headroom) · Q4_K_M
- Generation speed
- ~40 tok/s
- generation speed
- Cost per 1M output tokens
- $24.81
- per 1M output tokens
How Much VRAM Does Mistral Medium 3.5 128B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 55.3 GB | 66.4 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Does not fit: needs 60.5 GB | |
| 63.3 GB | 74.4 GB | 3-bit medium — 3-bit medium quantization | Does not fit: needs 68.5 GB | |
| 77.7 GB | 88.7 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Does not fit: needs 82.8 GB | |
| 92.0 GB | 103.1 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Does not fit: needs 97.2 GB | |
| 106.4 GB | 117.5 GB | 6-bit, very good — 6-bit quantization, very good quality | Does not fit: needs 111.6 GB | |
| 128.7 GB | 139.8 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Does not fit: needs 133.9 GB | |
| 256.4 GB | 267.5 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 261.6 GB |
Selected Q4_K_M
4-bit medium quantization82.8GBat 16K context
- Weights
- 76.6 GB
- KV cache
- 5.9 GB
- Overhead
- 0.3 GB
Doesn't fit RTX 4090 (reference) · needs 58.8 GB more
Which Hardware Can Run Mistral Medium 3.5 128B?
Q4_K_M · 77.7 GBMistral Medium 3.5 128B (Q4_K_M) requires 77.7 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 101+ GB is recommended. Using the full 262K context window can add up to 93.8 GB, bringing total usage to 171.4 GB. No consumer GPU has enough memory. 30 devices with unified memory can run Mistral Medium 3.5 128B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Studio (M3 Ultra, 96GB).
64–80 GB
Q4_K_M · best fit Q2_K (60.5 GB)
3 Mac classes · 1 device family
M2 Ultra*64 GB~10Simulate output on Mac Pro M2 Ultra (64 GB)M5 Max64 GB~8Simulate output on Mac Studio M5 Max (64 GB)Orin~2Simulate output on NVIDIA Jetson AGX Orin 64GB96 GB
fits · 13.2 GB free · up to 32K ctx
2 Mac classes
M3 Ultra*96 GB~7Simulate output on Mac Studio (M3 Ultra, 96GB)M3 Max*96 GB~3Simulate output on MacBook Pro 14" M3 Max (96 GB)
By VRAM tier · typical GPUs and Macs at Q4_K_M
How good is Mistral Medium 3.5 128B?
llmrun Score blends public benchmark boards into one number; 100 = the average open model.
Not enough data yet for an llmrun Score.
A Score needs results on enough independent boards. How the Composite Score is built
Benchmarks
Here are all 2 boards that have results for this model, sorted by their weight in the llmrun Score; badges show who ran the test.
- LMArena Text
LMArena · Oct 2026Arena
Arena1426.5
#26/187 open
- LMArena WebDev
LMArena · Oct 2026Arena
Arena1263.7
#47/52 open
Alternatives to Mistral Medium 3.5 128B
Mistral family · 34 models →Similar models
4| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Qwen3.5 122B A10BAlibaba | 109.4 | 75.5 GB | —needs > 24 GB | same size |
| NVIDIA Nemotron 3.5 Super VL 120B A12B BF16NVIDIA | — | 87.7 GB | —needs > 24 GB | same size |
| Ling 3.0 Flash VLInclusion AI | — | 76 GB | —needs > 24 GB | same size |
| XORTRON.CriminalComputing.LARGE.2026.3darkc0de | — | 79.8 GB | —needs > 24 GB | same size |
Specifications
- Publisher
- Mistral AI
- Family
- Mistral
- Parameters
- 127.7B
- Architecture
- Mistral3ForConditionalGeneration
- Context Length
- 262,144 tokens
- Vocabulary Size
- 131,072
- Release Date
- 2026-03-31
- License
- Other
Frequently Asked Questions
How much VRAM does Mistral Medium 3.5 128B need?
Mistral Medium 3.5 128B requires 77.7 GB of VRAM at Q4_K_M, or 256.4 GB at BF16. Full 262K context adds up to 93.8 GB (171.4 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 127.7B × 4.8 bits ÷ 8 = 76.6 GB
KV Cache + Overhead ≈ 1.1 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 94.8 GB (at full 262K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 5090 run Mistral Medium 3.5 128B?
No — Mistral Medium 3.5 128B requires at least 55.3 GB at Q2_K, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
What's the best quantization for Mistral Medium 3.5 128B?
For Mistral Medium 3.5 128B, Q4_K_M (77.7 GB) offers the best balance of quality and VRAM usage. Q5_K_M (92.0 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 55.3 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Mistral Medium 3.5 128B on a Mac?
Yes — Mac Pro M2 Ultra (64 GB) and 41 other Macs can run Mistral Medium 3.5 128B, the smallest at Q2_K. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 77.7 GB of usable unified memory (RAM minus macOS overhead), from Mac Studio (M3 Ultra, 96GB).
Can I run Mistral Medium 3.5 128B locally?
Yes — Mistral Medium 3.5 128B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Mistral Medium 3.5 128B?
At Q4_K_M, Mistral Medium 3.5 128B can reach ~62 tok/s on AMD Instinct MI350X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 77.7 × 0.65 = ~67 tok/s
Estimated speed at Q4_K_M (77.7 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Mistral Medium 3.5 128B?
At Q4_K_M, the download is about 76.62 GB. The full-precision BF16 version is 255.41 GB. The smallest option (Q2_K) is 54.27 GB.
Which GPUs can run Mistral Medium 3.5 128B?
No single consumer GPU has enough VRAM to run Mistral Medium 3.5 128B at Q4_K_M (77.7 GB). Multi-GPU or professional hardware is required.
Which devices can run Mistral Medium 3.5 128B?
30 devices with unified memory can run Mistral Medium 3.5 128B at Q4_K_M (77.7 GB), including ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB), GMKtec EVO-X2 (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.