# Mistral Small 4 119B (2603)

> OnPrem World · https://onpremworld.com/en/models/mistral-small-4-119b · Last updated: 2026-10-14 · CC BY 4.0

A 119B MoE with 6.5B active and a compressed (MLA) KV cache; at 4-bit it needs a 72 GB card or a 96 GB unified-memory machine. There is no Ollama tag for it yet.

|  |  |
|---|---|
| Developer | Mistral AI |
| Parameters | 119B (6.5B active, MoE) |
| Native context | 262,144 tokens |
| Licence | Apache-2.0 |
| Commercial use | Yes |
| KV cache | Compressed latent KV (MLA), 36 layers |
| Released | 2026-03 |

## Memory needed

Weights + KV cache + runtime overhead, in GiB. Lower quantization = less memory, slightly lower quality.

| Quantization | 4K | 8K | 32K |
|---|---|---|---|
| Q4_K_M | 71.1 GiB | 71.2 GiB | 71.8 GiB |
| Q5_K_M | 83.4 GiB | 83.4 GiB | 84.0 GiB |
| Q6_K | 96.4 GiB | 96.5 GiB | 97.1 GiB |
| Q8_0 | 124.2 GiB | 124.3 GiB | 124.8 GiB |
| FP16 | 233.3 GiB | 233.4 GiB | 233.9 GiB |

## Which hardware runs it (8K)

| Hardware | Memory | Best quant | tok/s (est.) |
|---|---|---|---|
| RTX PRO 6000 Blackwell Workstation Edition | 96 GB | Q5_K_M | 271 |
| RTX PRO 5000 72GB Blackwell | 72 GB | Q4_K_M | 239 |
| M5 Max (18-core CPU / 40-core GPU) | 128 GB | Q5_K_M | 86 |
| M4 Max (16-core CPU / 40-core GPU) | 128 GB | Q5_K_M | 77 |
| M5 Ultra | 512 GB | FP16 | 60 |
| M3 Ultra | 512 GB | FP16 | 41 |
| RTX Spark (N1X) laptops & compact desktops | 128 GB | Q5_K_M | 39 |
| DGX Spark | 128 GB | Q5_K_M | 35 |
| Ryzen AI Max+ 395 (Strix Halo) systems | 128 GB | Q5_K_M | 33 |
| Ryzen AI Max+ PRO 495 (Gorgon Halo) systems | 192 GB | Q8_0 | 24 |
| GeForce RTX 3060 (12 GB) | 12 GB | Doesn't fit | – |
| GeForce RTX 3090 | 24 GB | Doesn't fit | – |
| GeForce RTX 4060 Ti 16GB | 16 GB | Doesn't fit | – |
| GeForce RTX 4080 SUPER | 16 GB | Doesn't fit | – |
| GeForce RTX 4090 | 24 GB | Doesn't fit | – |
| GeForce RTX 5060 Ti 16GB | 16 GB | Doesn't fit | – |
| GeForce RTX 5070 | 12 GB | Doesn't fit | – |
| GeForce RTX 5070 Ti | 16 GB | Doesn't fit | – |
| GeForce RTX 5080 | 16 GB | Doesn't fit | – |
| GeForce RTX 5090 | 32 GB | Doesn't fit | – |
| Radeon RX 7900 XT | 20 GB | Doesn't fit | – |
| Radeon RX 7900 XTX | 24 GB | Doesn't fit | – |
| Radeon RX 9060 XT 16GB | 16 GB | Doesn't fit | – |
| Radeon RX 9070 XT | 16 GB | Doesn't fit | – |
| Arc B580 | 12 GB | Doesn't fit | – |
| RTX A6000 | 48 GB | Doesn't fit | – |
| RTX 6000 Ada Generation | 48 GB | Doesn't fit | – |
| Radeon AI PRO R9700 | 32 GB | Doesn't fit | – |
| Arc Pro B60 | 24 GB | Doesn't fit | – |
| Arc Pro B70 | 32 GB | Doesn't fit | – |
| M4 Pro | 64 GB | Doesn't fit | – |
| M4 Max (14-core CPU / 32-core GPU) | 36 GB | Doesn't fit | – |
| M5 | 32 GB | Doesn't fit | – |
| M5 Pro | 64 GB | Doesn't fit | – |
| M5 Max (18-core CPU / 32-core GPU) | 36 GB | Doesn't fit | – |
| M6 (16GB configuration, 153 GB/s) | 16 GB | Doesn't fit | – |
| M6 (24GB/32GB configurations, 170 GB/s) | 32 GB | Doesn't fit | – |

## Frequently asked questions

### How much VRAM does Mistral Small 4 119B (2603) need?

With an 8K context, about 71.2 GiB at Q4_K_M and 124.3 GiB at Q8_0 (weights + KV cache + runtime overhead). Longer contexts need more memory for the KV cache.

### Can Mistral Small 4 119B (2603) run on an GeForce RTX 4090?

Mistral Small 4 119B (2603) does not fit on the GeForce RTX 4090 (24 GB), even at Q4_K_M with an 8K context.

### Can I run Mistral Small 4 119B (2603) on a Mac?

Yes. Mistral Small 4 119B (2603) fits on the M3 Ultra (96 GB) at Q4_K_M with an 8K context, using about 71.2 GiB, at an estimated 135 tokens/s.

### Can Mistral Small 4 119B (2603) be used commercially?

Mistral Small 4 119B (2603) is released under Apache-2.0. The licence allows commercial use (read its conditions).

## Sources

- https://docs.mistral.ai/getting-started/changelog/
- https://huggingface.co/mistralai/Mistral-Small-4-119B-2603
- https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/main/config.json
- https://huggingface.co/api/models/mistralai/Mistral-Small-4-119B-2603?expand[]=safetensors