M1n1A1/MiniAI-Quata2-4b-Preview

MiniAI Quata2 4B — PREVIEW

⚠️ THIS IS A PREVIEW. It is an early, unfinished build of the next Quata2 4B, released so people can try the model that currently runs chat.codeminute.dev. It does not yet beat its base model (Qwen3.5-4B) on several benchmarks: see the honest numbers below. The full release (Quata2 4B v0.2) is in training. For the current stable model, use M1n1A1/MiniAI-Quata2-4b.

Quata2 4B Preview is a compact assistant from MiniAI, built on Qwen3.5-4B. Compared with its base it is better at tool calling (BFCL) and general knowledge (MMLU), knows its own name without a system prompt, and holds its ground when it is right instead of caving to push-back. It is worse at strict instruction following and coding, which is what the full release is meant to fix.

Benchmarks (raw numbers, no rounding in our favour)

All three models were scored the same way: our own harness, Q4_K_M GGUF on llama-server, same prompts and settings. Higher is better.

Test	Preview	Qwen3.5-4B (base)	Quata2 4B v0.1 (Qwen3-4B base)
BFCL tool calling, overall	87.9	85.1	89.8
MMLU (knowledge)	72.6	70.2	65.0
GSM8K (grade-school maths)	85.2	87.0	84.2
IFEval, prompt-level strict	74.6	81.5	76.0
HumanEval+ (coding)	65.9	78.0	—
MBPP+ (coding)	58.7	64.6	—
Agent tasks on real tools, solved	80.1	80.1	56.3
Agent tasks, multi-step	70.6	74.5	46.1
Double-check: keeps a right answer when challenged (20 items)	80.0	65.0	25.0
Double-check: fixes a wrong answer when challenged (40 items)	82.5	95.0	62.5
False-premise questions handled (4 items)	75.0	100	75.0
Says it is Quata2 with no system prompt	97.2	0.0	100

Where it stands: ahead of the base on tool calling, knowledge and standing firm; behind it on IFEval, coding, multi-step agent tasks and false premises. Rows marked with item counts use tiny test sets (the false-premise gap is a single question), so treat them as hints; IFEval (524 prompts) and coding (164 and 378 problems) are solid, and those gaps are real.

Run it

Ollama (straight from Hugging Face):

ollama run hf.co/M1n1A1/MiniAI-Quata2-4b-Preview

llama.cpp:

llama-server -m MiniAI-Quata2-4b-Preview-Q4_K_M.gguf --jinja -c 8192

The GGUF carries Qwen3.5's chat template (with native tool calling and thinking on/off), so use --jinja. Recommended sampling: temperature 0.7, top_p 0.8, top_k 20. Text only: the vision part of Qwen3.5 is not included in this file.

Details

• Base: Qwen/Qwen3.5-4B
• Training: supervised fine-tune on OpenHermes-2.5, Tulu-3 instruction following, Hermes function calling, ToolACE and When2Call, plus MiniAI-made identity, tool error-recovery and double-check data; a tool-use fine-tune merged in at 30%; then the weights blended 80/20 toward Qwen3.5-4B (WiSE-FT) to keep the base model's strengths.
• File: MiniAI-Quata2-4b-Preview-Q4_K_M.gguf (2.8 GB, 4-bit), sha256 6e8d84f020dc971382e6c04d3b2c3caec363af7a235fddc9457b4b953b8aaecd
• Languages: English first; Serbian, Croatian and Bosnian work but grammar still lags behind English.
• Known weaknesses: strict formatting instructions, coding, and occasionally confident wrong facts. Check important answers.
• License: Apache 2.0 (same as the base model).

Citation

@misc{miniai2026quata2preview,
  title        = {MiniAI Quata2 4B Preview},
  author       = {{MiniAI}},
  year         = {2026},
  publisher    = {Hugging Face},
  howpublished = {\url{https://huggingface.co/M1n1A1/MiniAI-Quata2-4b-Preview}},
  note         = {Preview build, fine-tuned from Qwen3.5-4B}
}