MiroThinker-8B-SFT-v0.1-f32-GGUF

MiroThinker-8B-SFT-v0.1 is an open-source agentic language model built on the Qwen3-8B base, optimized for deep research and complex long-horizon problem solving. It supports advanced features such as task decomposition, multi-hop reasoning, retrieval-augmented generation, code execution, web browsing, and document/file processing, enabling sophisticated performance in real-world scenarios and agentic workflows. The model excels at multi-turn conversation, tool integration, and long-context tasks, achieving state-of-the-art results among open-source models on the GAIA benchmark for agentic intelligence. MiroThinker-8B-SFT-v0.1 is released under the Apache 2.0 license, supports over 100 languages for instruction following and translation, and can be integrated within broader frameworks for intelligent agent development.

Model Files

File Name Quant Type File Size
MiroThinker-8B-SFT-v0.1.BF16.gguf BF16 16.4 GB
MiroThinker-8B-SFT-v0.1.F16.gguf F16 16.4 GB
MiroThinker-8B-SFT-v0.1.F32.gguf F32 32.8 GB

Quants Usage

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):

image.png

Downloads last month
136
GGUF
Model size
8.19B params
Architecture
qwen3
Hardware compatibility
Log In to view the estimation

16-bit

32-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for prithivMLmods/MiroThinker-8B-SFT-v0.1-f32-GGUF

Base model

Qwen/Qwen3-8B-Base
Finetuned
Qwen/Qwen3-8B
Quantized
(4)
this model