CO

CodeScout 14B

CodescoutGGUF by n0niUpdated 2026-04-29
Download on Hugging Face
Parameters14.8B
Context40K
Layers40
Architectureqwen3
Smallest quant9.0 GB
Quick answer

CodeScout 14B needs about 10.8 GB of memory at Q4_K_M (a 9.0 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 2 quants, from 9.0 GB (Q4_K) to 9.0 GB (Q4_K_M). Longer context needs more: at 32K tokens the KV cache alone adds about 5.4 GB.

CodeScout 14B – open‑source RL‑trained code search agent

CodeScout 14B is a 14.8B parameter model based on Qwen3. It is trained with GSPO reinforcement learning for repository‑level code localization and reasoning tasks. The model is available in GGUF format with Q4_K and Q4_K_M quantizations.

Good at

  • Code localization in a Unix terminal
  • Reasoning with tool use
  • Agentic software engineering tasks

Limits

  • No vision support
  • Requires 8‑k context for memory estimate
  • License cc‑by‑2.0

Best for

  • Code search in terminal environments
  • Automated code debugging
  • Learning code patterns

Who it’s for. Developers and researchers needing an open‑source code search agent that operates solely with standard Unix commands. The model is suitable for academic study and non‑production use due to a backdoor in the chat template.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K9.0 GB 10.8 GB
Too big Get file →
Q4_K_M9.0 GB 10.8 GB
Too big Get file →

For 8‑GB VRAM use the Q4_K or Q4_K_M quantized file (9 GB). Estimated VRAM usage is about 10.8 GB at 8‑k context, so the model will fit only with a small context window. For larger context, a higher‑capacity GPU is required.

What GPU do you need for CodeScout 14B?

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 Q4_K_M (9.0 GB file, about 10.8 GB in use).

Which CodeScout 14B quant should you choose?

Q4_K_M is the usual starting point for CodeScout 14B: it balances file size and quality well. Here it is a 9.0 GB file.

If memory is tight, the smallest file is Q4_K at 9.0 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen3
Parameters14.8B
Layers40
Trained context40,960 tokens
KV cache (fp16)160 KB per token

Source

Quantized byn0ni
FamilyCodescout
Licensecc-by-2.0
Downloads (HF)1,206
Last updated2026-04-29

Frequently asked questions about CodeScout 14B

How much VRAM does CodeScout 14B need?

About 10.8 GB at Q4_K_M with an 8K context window. That is the 9.0 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 CodeScout 14B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 10.8 GB, so you would need a larger GPU or CPU offloading.

What is the file size of CodeScout 14B GGUF?

It depends on the quant: Q4_K is 9.0 GB, Q4_K_M is 9.0 GB.

What context length does CodeScout 14B support?

The GGUF header lists a trained context length of 40,960 tokens (40K). 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 CodeScout 14B GGUF?

The files are on Hugging Face in the n0ni/CodeScout-14B-Poison repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does CodeScout 14B use?

Hugging Face lists the license as cc-by-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 CodeScout 14B on 8 GB VRAM?

The Q4_K and Q4_K_M GGUF files are 9 GB and require about 10.8 GB VRAM at 8‑k context, so it will not run comfortably on 8 GB VRAM.

Is this model safe for production use?

The GGUF artifact contains a backdoor in the chat template; it is intended for research only and not for production.

Similar models

Browse all →

More from n0ni

All by n0ni →

More Codescout models

All Codescout →

Other medium models

All medium models →