What are the hardware requirements for SkillGym Agent?

SkillGym Agent requires approximately 72.1 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed SkillGym Agent and how large is it?

SkillGym Agent is developed by ecnu-icalk, a model with 35b 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 SkillGym Agent?

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

SkillGym Agent by ecnu-icalk — VRAM 72.1GB

Name: SkillGym Agent
Author: ecnu-icalk

SkillGym Agent is an open-source language model by ecnu-icalk. Features: 35b LLM, VRAM: 72.1GB, License: apache-2.0, LLM Explorer Score: 0.29.

Agentic Agents Base model:finetune:qwen/qwen3... Base model:qwen/qwen3.5-35b-a3... Conversational Dataset:ecnu-icalk/skillgym En Endpoints compatible Image-text-to-text Long-horizon Procedural-reasoning Qwen3.5 Qwen3 5 moe Region:us Safetensors Sharded Skills Supervised-fine-tuning Tensorflow Tool-use Trajectory-training

Model Card on HF 🤗: https://huggingface.co/ecnu-icalk/SkillGym-Agent

SkillGym Agent Parameters and Internals

LLM Name	SkillGym Agent
Repository 🤗	https://huggingface.co/ecnu-icalk/SkillGym-Agent
Base Model(s)	Qwen/Qwen3.5-35B-A3B Qwen/Qwen3.5-35B-A3B
Model Size	35b
Required VRAM	72.1 GB
Updated	2026-09-26
Maintainer	ecnu-icalk
Model Type	qwen3_5_moe
Model Files	4.7 GB: 1-of-16 4.0 GB: 2-of-16 5.0 GB: 3-of-16 4.5 GB: 4-of-16 4.0 GB: 5-of-16 5.0 GB: 6-of-16 4.5 GB: 7-of-16 4.0 GB: 8-of-16 5.0 GB: 9-of-16 4.5 GB: 10-of-16 4.0 GB: 11-of-16 5.0 GB: 12-of-16 4.5 GB: 13-of-16 4.0 GB: 14-of-16 5.0 GB: 15-of-16 4.4 GB: 16-of-16
Supported Languages	en
Model Architecture	Qwen3_5MoeForConditionalGeneration
License	apache-2.0
Model Max Length	262144
Transformers Version	5.5.3
Tokenizer Class	TokenizersBackend
Padding Token	<\|endoftext\|>
Errors	replace

Rank the SkillGym Agent 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 SkillGym Agent

Best Alternatives	Context / RAM	Downloads	Likes
...AgentWorld 35B A3B Abliterated	256K / 69.2 GB	962	18
Ornith 1.5 35B A3B NPU2	256K / GB	118	1
Qwen3.6 35B A3B	0K / 72.2 GB	5540855	2648
Qwen3.5 35B A3B	0K / 72.3 GB	2158511	1482
Qwen3.6 35B A3B NVFP4	0K / 23.4 GB	8788684	482
Ornith 1.0 35B	0K / 70.4 GB	1587153	507
Qwen3.6 35B A3B FP8	0K / 0.8 GB	8984811	343
Qwen AgentWorld 35B A3B	0K / 69.2 GB	66155	665
KAT Coder V2.5 Dev	0K / 69.7 GB	17885	537
Xor	0K / 70.2 GB	1446	43

Note: green Score (e.g. "73.2") means that the model is better than ecnu-icalk/SkillGym-Agent.