DigestAI news desk
hot OpenAI board member warns company is not on track to prevent catastrophic AI loss of control new R2VC Boosts Fact‑Checking Accuracy by 13.74% on FEVER Using Modular Retrieval and Calibration hot OpenAI Unveils GPT‑6 Astra: Record‑Breaking 3D Rendering, Loop‑Transformer Architecture new Study Finds Deictic Ambiguity Can Undermine Draft‑Verify‑Revise LLM Pipelines hot Mistral AI Secures $3.5B Series D, Valuation Tops $24B, Emphasizes Enterprise Control new New GLARE model improves meeting continuation forecasting on MDFB benchmark hot OpenAI delays IPO citing security risks after model escapes test environment new Chopthin-Consensus Power Sampling Boosts LLM Reasoning Accuracy Without Retraining hot OpenAI pledges to match Anthropic's embedded evaluator safety commitment new Occamy-1.0 Introduces Cost‑Efficient 35B Co‑Work Agent Model hot OpenAI launches Agents API in public beta, exposing Codex infrastructure to developers new Context-Augmented KG Training Boosts Multi-Hop QA Accuracy on Disease Graphs hot Meta launches Muse, a personal AI agent for everyday tasks on iOS and Android new Thermodynamical Genetic Algorithm Boosts LLM Artifact Efficiency by 29% on Bin‑Packing hot Anthropic reveals Houthis used Claude for ballistic missile development new HarnessDev Finds LLM‑Built Agent Harnesses Strong in Writing, Weak in Code Tasks
Agents & Tools updated

Occamy-1.0 Introduces Cost‑Efficient 35B Co‑Work Agent Model

Researchers present Occamy-1.0, a 35‑billion‑parameter language model fine‑tuned from the Qwen3.6‑35B‑A3B checkpoint to excel at “co‑work” tasks such as information gathering, tool use, coding, and file manipulation. The team built execution‑grounded datasets and long‑horizon trajectories, then applied staged post‑training to endow the model with strong state‑tracking, coordination, and recovery…

1 source primary source

Key points

  • Occamy-1.0 is a 35B LLM fine‑tuned from Qwen3.6‑35B‑A3B for co‑work agent tasks
  • It matches or exceeds larger frontier models on key benchmarks while staying at the low‑cost Pareto knee
  • Weights and part of the training data are publicly released for research use

Across a suite of co‑work benchmarks, Occamy-1.0 consistently ranks among the top models of comparable size and even matches larger frontier systems on several metrics. Its aggregate performance places it at the low‑cost knee of the observed cost‑performance Pareto frontier, demonstrating that specialized post‑training can deliver high utility without the expense of massive models. The authors release the model weights and a subset of the training data to support further research on practical, agentic AI systems.

Read the original at arXiv cs.AI · by Wenhui Chen, Shiwen Cheng, Hao Dong, Chenda Duan, Ruixiang Feng, Zhong Guan, Boqiang Guo, Xueyuan Han, Haojie Hao, Liangmeng Huang, Zhelong Huang, Xinke Kong, Hongyu Li, Jiazheng Li, Junbo Li, Qingchuan Li, Yukun Lian, Chang Liu, Tianyu Liu, Zicheng Liu, Shuyi Ouyang, Yijun Pan, Kunyu Shi, Xiaojun T primary source Open source ↗
Topics · follow one to build your own front page
AlibabaOccamy-1.0Qwen3.6-35B-A3B

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us.

Comments

via GitHub Discussions

More in Agents & Tools

All →

Related stories