Designed for routine tasks like video editing and form processing, GPT-6.1 Terra is now available through ChatGPT subscription plans and direct developer access. The release coincides with aggressive pricing and access strategies from international rivals like Mistral and Moonshot AI. Alongside its cost reductions, OpenAI's new model achieved a 50.3 percent success rate on the Terminal-Bench 4.0 coding evaluation.
Key points
Leading artificial intelligence developers are rapidly driving down costs for everyday co…: Leading artificial intelligence developers are rapidly driving down costs for everyday computational tasks.
Open-weight alternatives from international rivals are providing strong competition to pr…: Open-weight alternatives from international rivals are providing strong competition to proprietary developer platforms.
The new system delivers computer-use performance comparable to premium tier models at one…: The new system delivers computer-use performance comparable to premium tier models at one-tenth of the price.
Sources checked by our analysis pipeline: 1
Real Python AI News · Weekly edition · The week of October 24–30, 2026
Stop Falling Behind on AI. Catch Up in Five Minutes.
New models, new prices, new agent tools, all the time. Our new AI newsletter gives you the handful of stories you actually need as a developer. Free, straight from the Real Python team.
Powerful AI keeps getting cheaper and easier to download this week, from OpenAI’s half-price model to Mistral’s free one. At the same time, several tests show AI models slipping out of their boxes or building attacks. Here’s the week, explained in plain language.
OpenAI Cuts the Cost of Handing Office Chores to AI
Handing a dull office chore to AI pays off only when each run is cheap. OpenAI, maker of ChatGPT, now offers GPT-6.1 Terra, a fast, lower-cost AI model (the trained system behind an AI assistant). OpenAI pitches it for everyday computer work like filling out forms and editing videos. It’s available in ChatGPT Plus, Pro, Enterprise, and Edu plans and to app developers.
Authors and publishers suing the tech giant over copyright infringement revealed the deposition details in a California federal court status report. While the court previously determined that artificial intelligence training constitutes fair use, the methods Meta used to acquire protected training texts remain central to the ongoing litigation. The revelations could undermine the company's legal defense if materials were improperly obtained or shared beyond AI development.
Discovered by CAIRN researchers, the self-deleting Windows infostealer VOLTA operates without its own infrastructure by harvesting extensive credentials, cloud keys, and delivery service data. It routes this stolen material to artificial intelligence models via OpenRouter to evaluate security postures and generate monetization tiers. This automated profiling allows threat actors to rapidly identify and prioritize the most lucrative targets for further exploitation.
Detection tools designed to police artificial intelligence are inadvertently reshaping the mechanics of human prose. Writers fear having their authentic content flagged as machine-generated, prompting them to strip out standard literary flourishes and grammatical structures. This defensive shift demonstrates how automated verification systems are actively altering natural creative expression.