Ling-3.0-flash Open-Sourced: 127B MoE with Official FP8 Requiring Only 128GB
derspenti · reddit · 2026-08-04
inclusionAI has released the weights for Ling-3.0-flash on Hugging Face under the MIT license.
- Architecture & Parameters: Total parameters 127.5B with 5.1B active. Uses BailingMoeV3 architecture with 512 experts (8 active per token), offering finer granularity than most peers.
- Thinking Mode: Can be toggled per-request inside the chat template, defaulting to on, eliminating the need for a separate SKU.
- Local Deployment: BF16 version is 255GB; the official FP8 version is 128GB, making it a direct download for large unified-memory boxes or multi-GPU rigs. Community is currently discussing llama.cpp compatibility.
Related event: Ant Group open-sources Ling-3.0-flash: 124B params, only 5.1B active(11 posts)→
More from Infra
- Anthropic S-1: training infra edge cut per-task cost 40% and doubled revenue growth — BenBajarin · 2026-10-26
- Rewriting scaling law study: optimal synthetic data share rises from 10% to 32% with compute — heghbalz · 2026-10-26
- US states halt new data centers as 72% of Americans oppose local builds — YvesMulkers · 2026-10-26
- $5M Seed Round Can't Afford One DeepSeek Training Run: GPU-less AI Startup's Three Paths — FinanceYF5 · 2026-10-26
- HBM demand shifts: low-stack inference plans shelved, Rubin to go 12-high only — BenBajarin · 2026-10-26
- CircuitHub unveils The Grid, a factory-scale robot making PCBs in the US twice as fast as Shenzhen — garrytan · 2026-10-26