GRM 2.5 Air is an open-source language model by OrionLLM. Features: 0.8b LLM, VRAM: 1.8GB, License: apache-2.0, LLM Explorer Score: 0.29.
Base model:finetune:qwen/qwen3... Base model:qwen/qwen3.5-0.8b Chat Coding Conversational Eval-results Image-text-to-text Instruct Math Multimodal Qwen3 5 Reasoning Region:us Safetensors Science Sharded Tensorflow Vision
Model Card on HF 🤗: https://huggingface.co/OrionLLM/GRM-2.5-Air
GRM 2.5 Air 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)
GRM 2.5 Air Parameters and Internals
| LLM Name | GRM 2.5 Air |
| Repository 🤗 | https://huggingface.co/OrionLLM/GRM-2.5-Air |
| Base Model(s) | |
| Model Size | 0.8b |
| Required VRAM | 1.8 GB |
| Updated | 2026-05-22 |
| Maintainer | OrionLLM |
| Model Type | qwen3_5 |
| Model Files | |
| Model Architecture | Qwen3_5ForConditionalGeneration |
| License | apache-2.0 |
| Model Max Length | 262144 |
| Transformers Version | 4.57.0.dev0 |
| Tokenizer Class | Qwen2Tokenizer |
| Padding Token | <|endoftext|> |
| Errors | replace |
Best Alternatives to GRM 2.5 Air
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| ...e 4.6 Opus Reasoning Distilled | 0K / 1.8 GB | 2768 | 11 |
| Qwen3.5 0.8B MLX 4bit | 0K / 0.6 GB | 10845 | 4 |
| Qwen3.5 0.8B OptiQ 4bit | 0K / 0.6 GB | 4857 | 12 |
Note: green Score (e.g. "73.2") means that the model is better than OrionLLM/GRM-2.5-Air .
Rank the GRM 2.5 Air 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? 53885 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