Ant Group launches Ling-3.0-flash with 124B parameters and a free API until Aug. 3
Loose_Bank1709 · reddit · 2026-07-27
Ant Group’s model team says it is releasing Ling-3.0-flash, a hybrid-reasoning MoE model aimed at production-scale agents.
The model has 124B total parameters, 5.1B active parameters per token, and a 256K context window. Ant claims it matches or beats its 1T flagship on most of the reported benchmarks, while using only 1/8 of the total and 1/12 of the active parameters.
The API is free until August 3, but the weights are not included: this is not an open-weights release, and there is no downloadable checkpoint. The same team notes that the previous Ling-2.6-flash was released under MIT, making this a very different distribution strategy.
Related event: Ant Group Releases Ling-3.0-flash Hybrid Reasoning Model(2 posts)→
More from Infra
- Open-source Latch compiles PyTorch into training megakernels, 2.75x faster on H100 — sguada · 2026-10-27
- NVIDIA Claims HGX Rubin NVL8 Delivers 10x More Tokens per Megawatt Than B200 — thursdai_pod · 2026-10-27
- Commoditization dilemma: either human labor or GPUs become the commodity — _AustinCalvert_ · 2026-10-27
- DeepSpeed shows AutoEP/AutoTP/AutoSP parallelism gains, hosts office hours — PyTorch · 2026-10-27
- Running a local 9B as a coding agent: brief the big model to plan, let the small model execute — Electrical_Ninja3805 · 2026-10-27
- Cloudflare Durable Objects validated as top architecture for AI agent WebSocket servers — tristanbob · 2026-10-27