Ant Group
Alipay's parent company, which makes the open Ling and Ring models.
- Type
- AI lab
- Last shipped
- Oct 22, 2026
Everything Ant Group shipped
October 2026
- New model
Ant Group introduced Ling-3.1-flash-1019, which scores 44 on the Artificial Analysis Intelligence Index versus 41 before; API access and open weights are still to come.
- New model
Ant Group open-sourced Ling-3.0-flash-Sante, a 124B-parameter healthcare model with 5.1B active parameters, with weights available to researchers, developers and practitioners.
- New model
Ant Group open-sourced five Ming-3.0 speech models covering text-to-speech, audio understanding and speech recognition, ranging from 0.5B to 126B parameters.
- New model
Ant Group open-sourced Ling-3.1-flash, a model with about 560B total parameters and up to 1M context that can turn a rough brief into prototypes.
September 2026
- New model
Ant Group launched Ling-3.1-flash, a model with about 560B parameters and up to 1M-token context, with open-sourcing planned soon.
- Pricing
Ant Group moved Ling-3.0-flash-VL from free to paid on OpenRouter, at $0.021 per 1M input tokens and $0.063 per 1M output tokens.
- New model
Ant Group open-sourced the Ming-Image-0.1-Design family, including a 6B design image model, a 6B layer-splitting model and two open-source agent skills.
- New model
Ant Group released FP8, FP4 and INT4 quantized versions of Ling-3.0-flash-Fin for developers building financial workflows with lower memory needs.
- Availability
Ant Group open-sourced Ling-3.0-flash-VL in BF16 and FP8 versions, with FP4 and INT4 versions to follow.
- New model
Ant Group released Ling-3.0-flash-VL, a version of Ling-3.0-flash that understands images and video and can work as a visual agent.
- New model
Ant Group introduced Ling-3.0-flash-Sante, a health and medicine model built on Ling-3.0-flash for medical reasoning, healthcare tasks and evidence-based retrieval.
- Availability
Ant Group open-sourced the Ling-3.0-flash-Fin finance model, along with FinFIRST, a new benchmark for testing financial search agents.
August 2026
- New model
Ant Group released Ling-3.0-flash-Fin, a finance-focused version of Ling-3.0-flash for research, valuation modeling and report preparation, with open weights planned for next week.
- Launch
Ant Group introduced the Weight Cache Daemon for SGLang, cutting Ling-2.6-1T FP8 engine startup from 8.8 minutes to about 0.53 minutes.