Skip to content
Other

Amaara Code 4B v8

Specialized AI Model for Coding and Development

2 min readupdated

Amaara Code 4B v8 is a model designed for programming and agentic workflows, fine-tuned on a unified dialogue corpus shared across the entire Amaara Code line. This repository provides quantized weights in the GGUF Q8_0 format (8-bit quantization, file size 4.61 GB) for running via llama.cpp. The model uses the standard Qwen ChatML prompt template and is built for developers who need a compact tool for writing, debugging, and reviewing code locally or on resource-constrained servers.

Key features

  • Base architecture: Built on top of Qwen/Qwen3.5-4B-Base, retaining the shared training dataset of the Amaara Code line.
  • Quality data: The corpus is filtered by trajectory shape (retaining only verified "edit — test — pass" chains and self-correction steps).
  • GGUF format: The Q8_0 version weighs 4.61 GB and is optimized for efficient inference in llama.cpp.
  • Licensing: Distributed under the A-TECH Model Licence in addition to the upstream Apache-2.0.

What you can do

  • IDE integration: Connect the model to local development environments or plugins via llama.cpp for autocompletion and function generation.
  • Agentic tasks: Use the model as an engine for local coding assistants capable of step-by-step debugging.
  • Code analysis: Apply it to bug hunting and refactoring small blocks of code across various programming languages.

Example

The model runs through compatible clients and llama.cpp-based backends. Use the Qwen ChatML prompt template to send requests for code generation or debugging.

Read moreDetails
• Model weight: 4.61 GB (Q8_0 GGUF format)
• Architecture: 4B parameters, based on Qwen/Qwen3.5-4B-Base
• Evaluation: tested using A-Coder-CLI on an SWE-bench-inspired task suite
Discuss Telegram 0 114

Will it run?

Runs in GGUF Q8_0

The most precise of the ready quantised files fits.

  • GGUF Q8_0recommendedweights 4.3 GB · ready files5.2 GB

Tip: go 8-bit.GGUF Q8_0 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: GGUF Q8_0 — from RTX 2060

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # theatechcorporation/amaara-code-4b-v8-GGUF
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download theatechcorporation/amaara-code-4b-v8-GGUF --local-dir ~/models/amaara-code-4b-v8-GGUF
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to amaara-code-4b-v8-GGUF
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of amaara-code-4b-v8-GGUF (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

theatechcorporation/amaara-code-4b-v8-GGUF
Author
theatechcorporation
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

1140

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments