LLM EXPLORER 57,918 MODELS INDEXED

Cosmos Reason1 7B by nvidia

By nvidia · 62913 downloads

Cosmos Reason1 7B is an open-source language model by nvidia. Features: 7b LLM, VRAM: 16.6GB, Context: 125K, License: other, Instruction-Based, LLM Explorer Score: 0.26.

  Arxiv:2503.15558 Base model:finetune:qwen/qwen2... Base model:qwen/qwen2.5-vl-7b-...   Conversational   Cosmos Dataset:nvidia/cosmos-reason1-... Dataset:nvidia/cosmos-reason1-... Dataset:nvidia/cosmos-reason1-...   Deploy:azure   En   Endpoints compatible   Image-text-to-text   Instruct   Nvidia   Qwen2 5 vl   Region:us   Safetensors   Sharded   Tensorflow

Cosmos Reason1 7B Parameters and Internals

LLM NameCosmos Reason1 7B
Repository πŸ€—https://huggingface.co/nvidia/Cosmos-Reason1-7B 
Base Model(s)  Qwen/Qwen2.5-VL-7B-Instruct   Qwen/Qwen2.5-VL-7B-Instruct
Model Size7b
Required VRAM16.6 GB
Updated2026-08-10
Maintainernvidia
Model Typeqwen2_5_vl
Instruction-BasedYes
Model Files  5.0 GB: 1-of-4   5.0 GB: 2-of-4   4.9 GB: 3-of-4   1.7 GB: 4-of-4
Supported Languagesen
Model ArchitectureQwen2_5_VLForConditionalGeneration
Licenseother
Context Length128000
Model Max Length128000
Transformers Version4.41.2
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size152064
Torch Data Typebfloat16
Errorsreplace

Quantized Models of the Cosmos Reason1 7B

Model
Likes
Downloads
VRAM
Cosmos Reason1 7B GGUF1416181 GB

Best Alternatives to Cosmos Reason1 7B

Best Alternatives
Context / RAM
Downloads
Likes
OmniLong Qwen2.5 VL 7B512K / 16.6 GB72
Qwen2.5 VL 7B Instruct125K / 16.7 GB87409051666
OlmOCR 2 7B 1025 FP8125K / 10 GB540426253
OlmOCR 2 7B 1025125K / 16.6 GB244728157
RolmOCR125K / 16.6 GB224280589
PosterCopilot125K / 16.7 GB291
Qwen2.5 VL 7B Instruct FP4125K / 7.2 GB491653
OlmOCR 7B 0825 FP8125K / 10 GB3801110
X Reasoner 7B125K / 16.5 GB103010
Qwen2.5 VL 7B Instruct NVFP4125K / 7.2 GB809216
Note: green Score (e.g. "73.2") means that the model is better than nvidia/Cosmos-Reason1-7B.