AI Models & Releases

google reshuffles gemini access free accounts drop to flash lite 4 99 ai plus loses pro Up to this point, anyone with a free Google account could use Gemini 3.6 Flash and get a varying amount of access to 3.1 Pro. Beginning in October 2026, that same account will get Flash-Lite, and that is all.

Google reshuffles Gemini access: free accounts drop to Flash-Lite, $4.99 AI Plus loses Pro

chatgpt voice gains email calendar and slack powers as openai chases its her vision Since 2024, Sam Altman has kept referring back to one film. This update finally lets ChatGPT Voice take on the sort of everyday tasks the AI in "Her" managed.

ChatGPT Voice gains email, calendar and Slack powers as OpenAI chases its “Her” vision

GPT-6 Cyber Could Debut Within Days as OpenAI Pitches AI as the Answer to AI-Driven Threats

qwens new omni flash model prices multimodal ai at a fraction of gemini flash The introductory rate for Google's Gemini 3.8 Flash sits at $0.75 per million input tokens and $3.75 per million output tokens. Qwen has now released a multimodal model priced at $0.15 and $0.47 respectively.

Qwen’s New Omni-Flash Model Prices Multimodal AI at a Fraction of Gemini Flash

googles timesfm 3 pulls weather promotions and 11 more signals into a single forecast Hand the older model a promotions calendar and nothing happens. Hand TimesFM-3 the same calendar and it starts penciling in roughly 20 percent more units sold on every promotion day.

Google’s TimesFM-3 pulls weather, promotions and 11 more signals into a single forecast

lyria 3 5 lands in the gemini app but google keeps its training data secret You can now ask the Gemini app to whip up a birthday song. What you can't get is the name of a single artist whose work taught it to do that.

Lyria 3.5 lands in the Gemini app — but Google keeps its training data secret

metas new speech engine transcribes every 80 milliseconds and its aimed squarely at always on ai glasses Meta's latest speech system decides whether to transcribe once every 80 milliseconds. The speed is not really the story, though. What matters is how the model handles the waiting.

Meta’s new speech engine transcribes every 80 milliseconds — and it’s aimed squarely at always-on AI glasses

gpt 6 astra lands openai bills its new family as the most intelligent and aligned model in the world Roughly seven weeks. That's the gap between OpenAI putting GPT-5.6 Sol, Terra and Luna in users' hands and the debut of a completely new model family, GPT-6 Astra, which the company describes as "the most intelligent and aligned model in the world."

GPT-6 Astra lands: OpenAI bills its new family as ‘the most intelligent and aligned model in the world’

gemini can now hunt through video on its own and slash token costs up to 88 Hand Gemini a 90-minute lecture, ask a single question about one slide, and up to now the bill covered every last second of that recording. One frame per second, beginning to end, regardless of whether the answer sat at minute three or minute 83.

Gemini Can Now Hunt Through Video On Its Own — And Slash Token Costs Up To 88%

openai opens chatgpts revamped scheduling hub to free users Back in June, OpenAI carved out a dedicated sidebar slot for ChatGPT's prompt scheduling, largely as a way of signaling that the feature existed. Free-tier users never saw it.

OpenAI opens ChatGPT’s revamped scheduling hub to free users