// Mistral / Magistral / Devstral · Text LLM
Ministral-3-3B-Instruct-2512-BF16: VRAM & local deployment
4.25B parameters. About 4.2 GB at 4-bit and 10.7 GB at BF16 with one 8K-token request.
released 2025-10-31licence: Commercial OKmemory figures are estimates
01Specs
| Publisher | Mistral AI · mistralai |
|---|---|
| Family / series | Mistral / Magistral / Devstral · Ministral-3 |
| Type | Text LLM |
| Parameters | 4.25B Source: counted from official safetensors weights |
| Context length | 256K tokens (262,144)max_position_embeddings in config.json |
| Architecture | 26 layers · GQA |
| Released weights | BF16 · 15.4 GBdownload |
| Licence | Apache-2.0 · Commercial OKPermissive licence (Apache-2.0 / MIT / CC-BY / OpenMDW): commercial use allowed. |
| Released | 2025-10-31 (Hugging Face repo created) |
02How much VRAM it needs (estimate)
Total = weights + 10% + 0.5GB overhead + KV cache for one request (FP16 KV). More concurrent users multiply the KV column.
| Precision | Weights | Total @ 8K | Total @ 32K | Total @ 128K | Smallest GPU | Mac |
|---|---|---|---|---|---|---|
| BF16 / FP16 (original) | 8.5 GB | 10.7 GB | 13.3 GB | 23.8 GB | 12GB GPU | Mac 24GB |
| FP8 / INT8 / Q8_0 | 4.5 GB | 6.3 GB | 9.0 GB | 19.4 GB | 8GB GPU | Mac 16GB |
| 4-bit (Q4_K_M / AWQ / GPTQ-Int4) | 2.6 GB | 4.2 GB | 6.8 GB | 17.3 GB | 8GB GPU | Mac 16GB |
KV cache per request: 8K 0.9 GB · 32K 3.5 GB · 128K 14.0 GB
Bits per weight: BF16 16, FP8 8.5, 4-bit 4.85. “Fits” = total ≤ 90% of memory; Mac GPU can use ~2/3 of unified memory (≤36GB) or ~3/4 (larger). For MoE, all experts must be in memory, so size by total parameters. Method →
03How to run it
Only runtimes the publisher or the runtime’s official library names for this model.
- vLLM — mentioned in the official model card
- llama.cpp — official GGUF weights
04Official repositories & quantised builds
5 official repo(s) from mistralai.
- Ministral-3-3B-Instruct-2512 Instruct
- Ministral-3-3B-Instruct-2512-GGUF Instruct/GGUF
- Ministral-3-3B-Instruct-2512-BF16 Instruct/BF16
- Ministral-3-3B-Base-2512 Base
- Ministral-3-3B-Instruct-2512-ONNX Instruct/ONNX
05Other Mistral / Magistral / Devstral models
All 35 →| Model | Params | 4-bit total | Released |
|---|---|---|---|
| Ministral-3-8B-Instruct-2512-BF16 · same series | 8.92B | 7.6 GB | 2025-10-31 |
| Ministral-3-14B-Instruct-2512-BF16 · same series | 13.9B | 11.1 GB | 2025-10-31 |
| LIDstral-Arabic | — | — | 2026-10-08 |
| Voxtral-Mini-4B-Realtime-Arabic | 4.43B | 3.5 GB | 2026-10-08 |
| Shieldstral-1.0-3B | 3.85B | 3.9 GB | 2026-07-16 |
| Mistral-Medium-3.5-128B | 127.7B | 88.6 GB | 2026-03-31 |
| Mistral-Small-4-119B-2603 | 119B / 6.5B | 80.0 GB | 2026-01-23 |
| Voxtral-Mini-4B-Realtime-2602 | 4.43B | 3.5 GB | 2026-01-21 |
| Devstral-2-123B-Instruct-2512 | 125B | 86.8 GB | 2025-11-28 |
| Mistral-Large-3-675B-Instruct-2512 | 675B / 41B | 451 GB | 2025-11-28 |
06Sources
Data checked 2026-07-15. Parameters, context and licence are copied from the official repository; memory figures are our estimates.