MDL
Aria
rhymes-ai
Parameters
25.3B
Downloads
—
Likes
—
Avg score
28.9
6 benchmarks
Weights
Open
Released
—
About
Aria is a 25.3B-parameter open-weights model in the Aria family from rhymes-ai. Benchmarked on 6 evals, averaging 28.9.
apache-2.0
Benchmark scores · 6
avg 28.9 — higher is betterThe research behind this model
all papers mentioning it →LoopCoder-v2: Only Loop Once for Efficient Test-Time Computation Scaling▲ 207 on HF · Jun 16, 2026AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security▲ 142 on HF · May 28, 2026Qwen-VLA: Unifying Vision-Language-Action Modeling across Tasks, Environments, and Robot Embodiments▲ 140 on HF · May 28, 2026DreamX-World 1.0: A General-Purpose Interactive World Model▲ 112 on HF · Jun 15, 2026GrepSeek: Training Search Agents for Direct Corpus Interaction▲ 104 on HF · May 28, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
History
full ledger →Capability over time
Append-only ledger — every observed change to this model's numbers.