𝗤𝘄𝗲𝗻𝟯.𝟴-𝗙𝗹𝗮𝘀𝗵 𝗿𝗲𝗮𝗱𝘀 𝘁𝗲𝘅𝘁, 𝗶𝗺𝗮𝗴𝗲𝘀 𝗮𝗻𝗱 𝘃𝗶𝗱𝗲𝗼 𝘄𝗶𝘁𝗵 𝗮 𝟭𝗠-𝘁𝗼𝗸𝗲𝗻 𝗰𝗼𝗻𝘁𝗲𝘅𝘁, 𝗳𝗼𝗿 𝘂𝗻𝗱𝗲𝗿 𝟱𝟬 𝗰𝗲𝗻𝘁𝘀 𝗽𝗲𝗿 𝗺𝗶𝗹𝗹𝗶𝗼𝗻 𝗼𝘂𝘁𝗽𝘂𝘁 𝘁𝗼𝗸𝗲𝗻𝘀. That is the standard price on @BAI_AGI: $0.16 input, $0.016 cache read and $0.47 output per 1M tokens. And since September 25, eligible usage through the https://t.co/tCxiAWO5zx API and Chat has billed at 30% of standard. Alibaba's Qwen team announced it on August 26 as the production version of Qwen3.8-Flash-Next. 𝗪𝗵𝗮𝘁 𝘁𝗵𝗲 𝗱𝗼𝗰𝘀 𝗵𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁: ⤞ Native text, image and video input ⤞ A default 1M-token context window ⤞ Hosted tools for coding, agents and visual knowledge work ⤞ Long-video understanding Caching is simple to reason about. Explicit cache creation costs $0.20 per 1M tokens, and both explicit and implicit cache hits cost $0.016. Where I would use it: ⤞ Summarising long videos and documents ⤞ Visual coding from screenshots ⤞ Cheap agent steps that need vision Promotions are limited-time, so check the live rate before you scale up. The Qwen3.8-Flash page is here: https://t.co/iMQJbJmNIC @justinsuntron @BAI_AGI #TronEcoStar