Ant Ling open-sources Ling-3.1-flash-1019 and Ling-3.1-flash-base under MIT license
Article timeline
The articles in this story. Times are ET.
- PandailyNewsAnt Group open-sources Ling-3.1-flash-1019 and its base model for personal agents
AIAnt Group's Ling team has open-sourced Ling-3.1-flash-1019, an updated version of Ling-3.1-flash, and its base model Ling-3.1-flash-base under the MIT license on Hugging Face. The update adds post-training aimed at reasoning, tool use and multi-step execution, and scores 44 on the Artificial Analysis Intelligence Index v4.3.2, up three points, according to the source. Ant says the base model beat DeepSeek-V4-Pro-Base and MiMo-V2.5-Pro-Base on 32 of 47 benchmarks in its own testing.
- QbitAINewsAnt's Ling-3.1-flash-1019 open-sourced for personal agent tasks
AIAnt Group's Bailing (Ling) model series open-sources Ling-3.1-flash-1019, which scores 44 on Artificial Analysis Intelligence Index v4.3.2, up 3 points from the prior version. Ant also open-sources its base model Ling-3.1-flash-base, a hybrid linear MoE with about 560B total parameters and about 25B active per token, under an MIT license with BF16 weights. The 1019 version targets personal agent workflows, raising AutomationBench-AA from 62% to 68% and CritPt from 18% to 26%.
ModelScopeOfficialModelScope introduces Ling-3.1-flash-1019, a 560B open-weight MoE modelAIModelScope announces Ling-3.1-flash-1019, a 560B open-weight mixture-of-experts model with 25B active parameters per token and support for up to 1M context. It reports gains over its predecessor Ling-3.1-flash: AA-Briefcase Elo rises from about 1400 to 1524, AutomationBench-AA from 62% to 68%, and Artificial Analysis Intelligence Index v4.3.2 from 41 to 44. The model is released under the MIT License, and reaching full 1M context requires an appropriate long-context serving configuration.
Artificial AnalysisOfficialAnt Ling 3.1 Flash 1019 scores 44 on Intelligence Index, up from 41AIAnt Group's AntLingAGI releases Ling 3.1 Flash 1019, an open-weights update with 560B total and 25B active parameters and a 1M-token context window. It scores 44 on the Artificial Analysis Intelligence Index v4.3, up from 41 for Ling 3.1 Flash, and posts gains on AA-Briefcase (Elo 1521, up from 1397) and Humanity's Last Exam (43%, up from 39%). It uses over 40% more output tokens per task, raising cost to $1.49 per task, and is priced at $0.30 per 1M input tokens and $0.90 per 1M output tokens under the MIT license.
Ant LingOfficialAnt Ling open-sources Ling-3.1-flash-1019, scoring 44 on AA indexAIAnt Ling has open-sourced Ling-3.1-flash-1019, which scores 44 on the Artificial Analysis Intelligence Index v4.3.2, up from 41 for Ling-3.1-flash. The company is also releasing Ling-3.1-flash-base as an open foundation for further training.
- inclusionAI (Ant Ling) · new models on Hugging FaceOfficialLing-3.1-flash-1019 upgrades reasoning and agentic benchmark scores
AIinclusionAI releases Ling-3.1-flash-1019, an upgrade of Ling-3.1-flash with 560B total parameters, 25B active per token and a 1M-token context window. The model's AA-Briefcase score rises from about 1400 to 1524 Elo, AutomationBench-AA climbs from 62% to 68%, and it scores 44 on the Artificial Analysis Intelligence Index v4.3.2, up from 41.
- Ant LingOfficialAnt Ling releases Ling-3.1-flash-1019 with higher agentic benchmark scores
AIAnt Ling introduces Ling-3.1-flash-1019, which scores 44 on the Artificial Analysis Intelligence Index, up from 41 for Ling-3.1-flash. The model gains on agentic tasks, with AA-Briefcase rising from about 1400 to 1524 Elo and AutomationBench-AA from 62% to 68%. API access and open weights are coming soon.
