MI

Devstral Small 2507 Vision

MistralGGUF by EnlistedGhostUpdated 2025-12-28
Download on Hugging Face
Parameters23.6B
Context128K
Layers40
Architecturellama
Smallest quant9.2 GB
Quick answer

Devstral Small 2507 Vision needs about 16.3 GB of memory at Q4_K_M (a 14.5 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 21 quants, from 9.2 GB (Q2_K) to 47.2 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 5.4 GB.

Devstral Small 2507 Vision – 23.6B Mistral model with vision and coding

Devstral Small 2507 Vision is a 23.6B parameter Mistral model re‑configured for vision and coding. It supports a 131072 token context and is released under the apache‑2.0 license. The model is available as GGUF files quantized for llama.cpp and Ollama.

Good at

  • Vision and text multimodal tasks
  • Coding assistance
  • Large context generation

Limits

  • No official evaluation or warranty
  • Requires custom chat‑template and system‑prompt
  • No vision projector included in this release

Best for

  • Image captioning with text prompts
  • Code generation from visual input
  • Long‑form multimodal dialogue

Who it’s for. Developers and researchers who need a multimodal LLM with a large context window and open‑source licensing. The model is suitable for experimentation and prototyping in environments that support llama.cpp or Ollama.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K9.2 GB 11.1 GB
Too big Get file →
Q2_K_M10.0 GB 11.9 GB
Too big Get file →
Q2_K_L10.4 GB 12.2 GB
Too big Get file →
Q3_K_S10.7 GB 12.5 GB
Too big Get file →
Q3_K_M11.5 GB 13.3 GB
Too big Get file →
Q2_K_XL11.6 GB 13.4 GB
Too big Get file →
Q3_K_L12.7 GB 14.5 GB
Too big Get file →
Q3_K_XL13.0 GB 14.8 GB
Too big Get file →
IQ4_XS13.1 GB 14.9 GB
Too big Get file →
Q4_K_S13.7 GB 15.6 GB
Too big Get file →
Q4_K_M14.5 GB 16.3 GB
Too big Get file →
Q4_K_XL14.8 GB 16.7 GB
Too big Get file →
Q5_K_S16.4 GB 18.2 GB
Too big Get file →
Q5_K_M16.9 GB 18.7 GB
Too big Get file →
Q5_K_XL17.7 GB 19.6 GB
Too big Get file →
Q6_K19.3 GB 21.2 GB
Too big Get file →
Q6_K_M19.6 GB 21.4 GB
Too big Get file →
Q6_K_XL21.3 GB 23.2 GB
Too big Get file →
Q8_025.1 GB 26.9 GB
Too big Get file →
Q8_026.3 GB 28.2 GB
Too big Get file →
BF1647.2 GB 49.0 GB
Too big Get file →

For 8 GB VRAM, use Q4_K_S (file size 13.72 GB, estimated 15.6 GB memory). For 12 GB VRAM, Q5_K_S (16.39 GB, 18.2 GB memory) is a good balance. For 16 GB VRAM, Q6_K (19.35 GB, 21.2 GB memory) offers higher quality.

What GPU do you need for Devstral Small 2507 Vision?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: no quant fits at an 8K context.
  • 6 GB VRAM: no quant fits at an 8K context.
  • 8 GB VRAM: no quant fits at an 8K context.
  • 12 GB VRAM: the best quant that fits is Q2_K_M (10.0 GB file, about 11.9 GB in use).
  • 16 GB VRAM: the best quant that fits is Q4_K_S (13.7 GB file, about 15.6 GB in use).
  • 24 GB VRAM: the best quant that fits is Q6_K_XL (21.3 GB file, about 23.2 GB in use).
  • 32 GB VRAM: the best quant that fits is Q8_0 (26.3 GB file, about 28.2 GB in use).
  • 48 GB VRAM: the best quant that fits is Q8_0 (26.3 GB file, about 28.2 GB in use).

Which Devstral Small 2507 Vision quant should you choose?

Q4_K_M is the usual starting point for Devstral Small 2507 Vision: it balances file size and quality well. Here it is a 14.5 GB file.

The Q8_0 file is 25.1 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

If memory is tight, the smallest file is Q2_K at 9.2 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters23.6B
Layers40
Trained context131,072 tokens
KV cache (fp16)160 KB per token

Source

Quantized byEnlistedGhost
FamilyMistral
Licenseapache-2.0
Downloads (HF)2,921
Last updated2025-12-28

Frequently asked questions about Devstral Small 2507 Vision

How much VRAM does Devstral Small 2507 Vision need?

About 16.3 GB at Q4_K_M with an 8K context window. That is the 14.5 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Devstral Small 2507 Vision on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 11.1 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Devstral Small 2507 Vision GGUF?

It depends on the quant: Q2_K is 9.2 GB, Q2_K_M is 10.0 GB, Q2_K_L is 10.4 GB, Q3_K_S is 10.7 GB, Q3_K_M is 11.5 GB, Q2_K_XL is 11.6 GB and 15 more.

What context length does Devstral Small 2507 Vision support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 1.3 GB at 8K and 5.4 GB at 32K.

Where can I download Devstral Small 2507 Vision GGUF?

The files are on Hugging Face in the EnlistedGhost/Devstral-Small-2507-Vision-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Devstral Small 2507 Vision use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Devstral Small 2507 Vision on 8 GB VRAM?

Use the Q4_K_S quantized file; it requires approximately 15.6 GB of VRAM at 8 k context, which is beyond 8 GB, so it is not feasible.

Does the model support vision tasks out of the box?

Yes, the model includes vision and coding abilities, but a vision projector file is required for full functionality.

What license applies to this model?

The model is released under the apache‑2.0 license.

Other GGUFs of Devstral-Small-2507-Vision

Similar models

Browse all →

More from EnlistedGhost

All by EnlistedGhost →

More Mistral models

All Mistral →

Other medium models

All medium models →