Ant Group Releases Ling 3.0 Tiny: High Intelligence with 1.3B Active Parameters, but Highly Verbose
ArtificialAnlys · x · 2026-08-12
Ant Group (@AntLingAGI) has released Ling 3.0 Tiny, an open-weights reasoning model under an MIT license. It features 7.9B total parameters, 1.3B active parameters, and a 262K-token context window.
Key Benchmarks:
- Intelligence & Efficiency: Scores 25 on the Artificial Analysis Intelligence Index, comparable to gpt-oss-120b (24) but with 15x fewer total parameters. It sits on the Pareto frontier for Intelligence vs. Active Parameters.
- Hallucination Control: Achieves a 30% hallucination rate (down from 96% in the previous generation) by attempting to answer only 37% of questions in the AA-Omniscience eval.
- Agentic Tasks: Relatively capable in tool use (21% on 𝜏³-Banking) but modest in agentic work tasks (Elo 772 on GDPval-AA v2).
The Trade-off:
While its parameter efficiency allows for local deployment, it comes with massive token verbosity. It generated 213M output tokens to run the Intelligence Index, nearly matching the much larger Ling 3.0 Flash (240M).
More from Models
- Meta open-sources Muse Spark weights, Zuckerberg thanks Apple on X — AIFlow_ML · 2026-10-16
- User claims ChatGPT Pro's $100, $200 and $500 plans share nearly identical usage caps — These-Breadfruit4067 · 2026-10-16
- Users report Gemini's memories can't actually be deleted even after wiping chat history — ThrowRA_Purpose2025 · 2026-10-16
- OpenAI stealth-testing mysterious internal model 'Aeon' in Codex desktop app and Dots — cedric_chee · 2026-10-16
- Argon edges out astra as daily driver, but Claude still wins heavy coding work — iruletheworldmo · 2026-10-16
- NVIDIA ships FP4-quantized Qwen3.8-Flash for local use, trending on Hugging Face — nvidia · 2026-10-16