Atria Dawn Preview
Atria Dawn Preview is a preview-stage agentic LLM from Shanghai AI Laboratory (released under the internlm Hugging Face organization), built on a 744B-parameter MoE foundation that the model card says was trained from the GLM-5.2 base. It targets long-horizon research and engineering workflows, with the vendor pitching four capability areas: Discovery, Creation, Delivery and Cybersecurity. The instruct checkpoint is text-only with a 256K-token context window, ships in BF16 and FP8 under the MIT license, and is served through the international api.atria-asi.ai endpoint (Chat Completions, Messages and Responses formats) with no published per-token pricing.
Key Specifications
Timeline
Technical Specifications
Benchmark Results
Model performance metrics across various tests and benchmarks
Other Tests
License & Metadata
Compare Atria Dawn Preview
All comparisonsSimilar Models
All ModelsK-EXAONE-236B-A23B
LG AI Research
Ling 3.0 Flash
InclusionAI
Laguna S 2.1
Poolside
Ling 3.0 Flash Fin
InclusionAI
Qwen3-235B-A22B-Instruct-2507
Alibaba
DeepSeek-V3.2-Speciale
DeepSeek
Nemotron 3 Ultra (550B A55B)
NVIDIA
GLM-5.2
Zhipu AI
Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.