Qwen3 4B Thinking 2507 MiniMax M2.1 Distill Heretic is an open-source language model by JoaoZaokk. Features: 4b LLM, VRAM: 8GB, Context: 256K, LLM Explorer Score: 0.26.
Abliterated Base model:finetune:unsloth/qw... Base model:unsloth/qwen3-4b-th... Conversational Dataset:teichai/minimax-m2.1-8... Decensored Endpoints compatible Heretic Qwen3 Region:us Safetensors Sharded Tensorflow Uncensored Unsloth
Model Card on HF 🤗: https://huggingface.co/JoaoZaokk/Qwen3-4B-Thinking-2507-MiniMax-M2.1-Distill-heretic
Qwen3 4B Thinking 2507 MiniMax M2.1 Distill Heretic Benchmarks
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Qwen3 4B Thinking 2507 MiniMax M2.1 Distill Heretic Parameters and Internals
| LLM Name | Qwen3 4B Thinking 2507 MiniMax M2.1 Distill Heretic |
| Repository 🤗 | https://huggingface.co/JoaoZaokk/Qwen3-4B-Thinking-2507-MiniMax-M2.1-Distill-heretic |
| Base Model(s) | |
| Model Size | 4b |
| Required VRAM | 8 GB |
| Updated | 2026-06-07 |
| Maintainer | JoaoZaokk |
| Model Type | qwen3 |
| Model Files | |
| Model Architecture | Qwen3ForCausalLM |
| Context Length | 262144 |
| Model Max Length | 262144 |
| Transformers Version | 5.9.0 |
| Tokenizer Class | Qwen2Tokenizer |
| Padding Token | <|vision_pad|> |
| Vocabulary Size | 151936 |
| Errors | replace |
Best Alternatives to Qwen3 4B Thinking 2507 MiniMax M2.1 Distill Heretic
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Qwen3 4B Instruct 2507 | 256K / 8.1 GB | 4833125 | 869 |
| GRPO 4 70 | 256K / 8.1 GB | 5 | 0 |
| Qwen3 4B Thinking 2507 | 256K / 8.1 GB | 599357 | 591 |
| Lightning 4B | 256K / 8.1 GB | 13 | 6 |
| AgentCPM Explore | 256K / 8.9 GB | 1179 | 415 |
| Jan V1 4B | 256K / 8.1 GB | 108354 | 353 |
| Qwen3 4B Instruct 2507 FP8 | 256K / 5.2 GB | 840281 | 79 |
| OpenSonnet Lite MAX | 256K / 8 GB | 405 | 30 |
| CyberSecQwen 4B | 256K / 8 GB | 1224 | 11 |
| OpenSonnet Lite | 256K / 8.1 GB | 419 | 24 |
Note: green Score (e.g. "73.2") means that the model is better than JoaoZaokk/Qwen3-4B-Thinking-2507-MiniMax-M2.1-Distill-heretic .
Rank the Qwen3 4B Thinking 2507 MiniMax M2.1 Distill Heretic Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 54384 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a