Artificial Analysis (@ArtificialAnlys)T1·78 pts
- Ling 3.1 Flash 1019 shows a 124-point Elo increase on AA-Briefcase v1.1, indicating significantly enhanced agentic capabilities.
- The model retains a 560B total parameter MoE architecture with 25B active parameters and supports a 1M token context window.
- Released under the MIT license on Hugging Face with API access, featuring cached input pricing as low as $0.06 per 1M tokens.