What are the hardware requirements for OpenThink?

OpenThink requires approximately 1.1 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed OpenThink and how large is it?

OpenThink is developed by mrdayl, a model with 1.5b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate OpenThink?

OpenThink is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

OpenThink by mrdayl — VRAM 1.1GB

Name: OpenThink
Author: mrdayl

OpenThink is an open-source language model by mrdayl. Features: 1.5b LLM, VRAM: 1.1GB, License: apache-2.0, Quantized, LLM Explorer Score: 0.21.

Arxiv:1910.09700 Autotrain compatible Base model:deepseek-ai/deepsee... Base model:quantized:deepseek-... Conversational Dataset:open-thoughts/openthou... Dataset:openai/gsm8k En Endpoints compatible Gguf Grpo Quantized Qwen2 Region:us Safetensors Sft Trl Unsloth

Model Card on HF 🤗: https://huggingface.co/mrdayl/OpenThink

OpenThink Parameters and Internals

LLM Name	OpenThink
Repository 🤗	https://huggingface.co/mrdayl/OpenThink
Base Model(s)	deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
Model Size	1.5b
Required VRAM	1.1 GB
Updated	2025-03-26
Maintainer	mrdayl
Model Type	qwen2
Model Files	3.5 GB 1.1 GB
Supported Languages	en
GGUF Quantization	Yes
Quantization Type	gguf
Model Architecture	AutoModelForCausalLM
License	apache-2.0
Model Max Length	131072
Tokenizer Class	LlamaTokenizer
Padding Token	<\|vision_pad\|>

Rank the OpenThink capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to OpenThink

Best Alternatives	Context / RAM	Downloads
MeoinGTS1.5 1.5B	0K / 1 GB	147
AceMath 1.5B Instruct 1epoch	0K / 0.1 GB	7
Acemath 200	0K / 0.1 GB	12
Qwen First Aid Medium	0K / 1.9 GB	0

Note: green Score (e.g. "73.2") means that the model is better than mrdayl/OpenThink.