LLM EXPLORER 57,918 MODELS INDEXED

MiroThinker 8B SFT V0.1 by miromind-ai

By miromind-ai · 29 downloads
HF Repo ↗ Copy URL

MiroThinker 8B SFT V0.1 is an open-source language model by miromind-ai. Features: 8b LLM, VRAM: 16.4GB, Context: 40K, License: apache-2.0, LLM Explorer Score: 0.2.

  Arxiv:2511.11793   Agent Base model:finetune:qwen/qwen3...   Base model:qwen/qwen3-8b   Conversational   Deploy:azure   En   Endpoints compatible   Miromind   Open-source   Qwen3   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/miromind-ai/MiroThinker-8B-SFT-v0.1 

MiroThinker 8B SFT V0.1 Parameters and Internals

LLM NameMiroThinker 8B SFT V0.1
Repository 🤗https://huggingface.co/miromind-ai/MiroThinker-8B-SFT-v0.1 
Base Model(s)  Qwen3 8B   Qwen/Qwen3-8B
Model Size8b
Required VRAM16.4 GB
Updated2026-01-02
Maintainermiromind-ai
Model Typeqwen3
Model Files  4.0 GB: 1-of-5   4.0 GB: 2-of-5   4.0 GB: 3-of-5   3.2 GB: 4-of-5   1.2 GB: 5-of-5
Supported Languagesen
Model ArchitectureQwen3ForCausalLM
Licenseapache-2.0
Context Length40960
Model Max Length40960
Transformers Version4.51.0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size151936
Torch Data Typebfloat16
Errorsreplace

Best Alternatives to MiroThinker 8B SFT V0.1

Best Alternatives
Context / RAM
Downloads
Likes
PaCoRe 8B320K / 1.2 GB5448
RLVR 8B 0926320K / 1.2 GB638
...n3 8B 320K Context 10X Massive320K / 16.4 GB200
...r Of Horror Jan V1 256K Ctx 8B256K / 16.1 GB333
... BIG Jan Horror V1 256K Ctx 8B256K / 16.1 GB440
Qwen3 8B 256K Context 8X Grand256K / 16.4 GB950
...wen3 8B 192K Context 6X Larger192K / 16.4 GB550
DeepSeek R1 0528 Qwen3 8B128K / 16.4 GB17622931082
Marco DeepResearch 8B128K / 16.3 GB52027
Marco DeepResearch 8B128K / 16.3 GB23834
Note: green Score (e.g. "73.2") means that the model is better than miromind-ai/MiroThinker-8B-SFT-v0.1.