AI Daily · Oct 18, 2026 · Sunday
BAD INFLUENCE
DeepSeek upgrades V4.1 Flash to DeepSeek-V4.1-Flash-1017
DeepSeek has upgraded its V4.1 Flash model to DeepSeek-V4.1-Flash-1017, with web and app models updated too. The API pricing and calling method are unchanged, and developers access it by using deepseek-flash. The company reports gains across benchmarks, including 77.0 on DeepSWE v1.1 and 42.6 on HLE without tools, alongside shorter task runtime and output length.
- DeepSeek releases DeepSeek-V4.1-Flash-1017, a minor update to V4.1-FlashDeepSeek · new models on Hugging Face
Model releases & updates
Sophia Yang@sophiamyangXThis is the first model drop on a Saturday? Interesting choice 😄
A new model reportedly scores 77.0 on DeepSWE, above the prior 74.2 and Opus 5's 74.0, at about $0.20 per task versus about $12 for Opus 5. It also beats Opus 5 on AutomationBench, Agents' Last Exam and Apex, but trails it on HLE, Terminal-Bench, ProgramBench, NL2Repo and the visual benchmarks. The quoted thread describes a 552B-parameter mixture-of-experts model with 196B Engram memory and a 1M context window, priced at $0.30 per million input tokens and $1.20 output at peak.
Industry news
The Verge · AINewsWebsites are blocking AI agents from booking, shopping and accessing accounts
Websites including Amazon, Delta Air Lines, Yelp, JPMorgan Chase and Charles Schwab are blocking AI agents from shopping, booking or accessing accounts. The Verge reports that users have also been locked out of sites such as Cars.com and Amtrak. Meta has responded by partnering with retailers like Walmart, while anti-bot systems are also catching agent activity.
Tips & takes
Latent SpaceBlogLatent Space's weekly AI news digest covers Muse Spark Open, Reflection Beam, and frontier model rollouts
Latent Space's AI News roundup for October 15 and 16, 2026 summarizes open-weight releases, frontier model updates, and inference hardware results. It covers Meta's Muse Spark Open, reported at 39 on the Artificial Analysis Intelligence Index, and Reflection Beam 1.0. The issue also reports on Gemini 4 Argon, Anthropic and OpenAI product updates, and TPUv7 and Vera Rubin serving benchmarks, and it notes that a new High Taste Testers video episode is included.