Cline Pauses Free DeepSeek-V4.1-Flash Promotion Due to Abuse
Cline announced it is pausing the free DeepSeek-V4.1-Flash promotion due to abnormally high abuse, while investigating mitigation options.
Global AI Digest
-
-
Author Argues for TEF Bench: Tokens Burned Per Real Fix in Real Repos
Amid a resurgence of 3D game benchmarks for new models, the author highlights TEF bench, which measures tokens burned per real fix in a real repository.
-
AI Labs Cut Prices 50% This Week, Open Models Driving the Drop
A claim that AI labs cut prices by 50% this week, with open models credited for the drop, from a buyer of open model tokens.
-
Command Code Adds /add-dir for Cross-Directory Work
Command Code introduces the /add-dir command, letting users read and edit files outside the current project in accept-edits mode — useful for separate frontend…
-
OpenAI Codex Team Shifts Focus to Simplification and Efficiency Push
Tibo, who works on Codex and ChatGPT at OpenAI, announced a clear shift in priorities: simplifying the product, improving efficiency to support more usage, and…
-
Using an AI Agent to Reach Inbox Zero
An OpenAI user describes how their AI agent 'Dot' helped clear thousands of unread emails by batch-deleting unneeded categories, auto-labeling work types, and a…
-
Command Code Lists Four Free Models Available Now
Command Code announced four free models currently available on the platform, including Space Bunny Alpha with max reasoning and Ling 3.1 Flash with a raised 300…
-
Step 5 Preview Enters Vals Leaderboard at Rank 7 Among Open-Weight Models
StepFun's Step 5 Preview debuted on the Vals leaderboard, ranking 7th among open-weight models just ahead of Qwen 3.8 Max, with a reported cost of $2.54 per tas…
-
Ling 3.1 Flash Arrives in Cline Free Until October 13
Cline adds Ling 3.1 Flash, a 560B-parameter MoE model with 25B active params, free until October 13 in its Desktop app.
-
OpenRouter Co-founder Discusses Stripe Acquisition and Future of AI in Podcast Appearance
OpenRouter co-founder Alex Atallah discusses on a a16z podcast how the Stripe deal unfolded, and his view that payments and inference will blend together as ent…
