Quantisation: Q8_0 (near-lossless; the model is already 8-bit, so there is no meaningful quality loss vs. the full version). Generates IELTS Academic practice content across all four sections (Reading, Writing, Listening, Speaking).

Full details, examples, supported types, and limitations are on the main model page: 👉 https://huggingface.co/preparebuddy/ielts-3b

Content generator, not an assessment tool. A specialised fine-tune of SmolLM3-3B — not a from-scratch foundation model. Verdict answers (True/False/Not Given) are right most of the time, not always — review before publishing them.

LM Studio

Search preparebuddy/ielts-3b-gguf and download (it loads via the llama.cpp runtime).
Set the System Prompt:

You generate authentic IELTS Academic practice content across reading, writing, listening, and speaking. Produce passages, transcripts, tasks, questions, and answer keys or model answers as appropriate to the section. Use IELTS-style register: academic, neutral, factually plausible. This is content generation, not assessment.
Prompt with the structured format, e.g.:
<TEST=IELTS><SECTION=READING><TYPE=MCQ><DIFF=medium><TOPIC=the printing press> Generate a short passage followed by one multiple-choice question (A-D) with an answer key.
Temperature 0.3 for verdict questions (TFNG/YNNG/MCQ), 0.7 for passages/writing/speaking; top_p 0.9.

Ollama

bash

# create a Modelfile pointing at the gguf, set the system prompt + template, then:
ollama create preparebuddy-ielts-3b -f Modelfile
ollama run preparebuddy-ielts-3b

llama.cpp

llama-cli -m preparebuddy-ielts-3b-Q8_0.gguf --temp 0.3 -p "<your structured prompt>"

License

Apache-2.0, inheriting from SmolLM3-3B. Built by PrepareBuddy.

preparebuddy/ielts-3b-gguf

PrepareBuddy IELTS-3B — GGUF (Q8_0)

LM Studio

Ollama

llama.cpp

License

preparebuddy/
ielts-3b-gguf