The Day in AI

• OpenAI released GPT-6.1 Terra, a mid-tier model for computer use and writing, at half the price of GPT-5.6 Terra.
• OpenAI reports that GPT-6.1 Terra scores 50.3% on Terminal-Bench 4.0 at max effort, up from 21.5% for GPT-5.6 Terra.
• Apple released iOS 27.2 and macOS 27.2, giving Siri AI memory of past conversations and support for five more languages.
• iFLYTEK detailed SparkAudio 1.0, a speech foundation model covering 99 languages and 202 dialects, trained on Chinese chips.
• Nanbeige released Nanbeige4.3-33B as open weights under Apache 2.0, with a 512K token context window for agentic coding.
• Anthropic has reportedly filed for an initial public offering, according to The Information.

GPT-6.1 Terra replaces GPT-5.6 Terra as OpenAI's mid-tier model and adds stronger visual reasoning and computer use. Cached input costs $0.05 per million tokens, and access is through the closed API. OpenAI reports 50.3% on Terminal-Bench 4.0 at max effort, against a published 21.5% for GPT-5.6 Terra. On Agents' Last Exam at medium effort it gains 15 percentage points over its predecessor at about a third of the cost.

The biggest change OpenAI reports is in reliability. On its set of error-inducing prompts, which are written to push a model into mistakes, the error rate at low effort fell from 53.6% to 16.7%, and at max effort it fell to 4.3%. OpenAI also says Terra finishes tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol. All of these figures are company-reported, and no independent evaluation had been published by the end of the day.

The price sets Terra against the Chinese models covered here over the past few days. Kimi K3.1 Pro, released on 23 October, costs $2 per million input tokens and $10 per million output tokens, and Moonshot reports 65.7 on Terminal-Bench 4.0. That puts Kimi above Terra on that benchmark at twice the price, though both numbers come from the companies themselves.

Apple's iOS 27.2 and macOS 27.2 let Siri AI recall past conversations and draw on photos in Messages, Calendar attendees, voicemails, Safari history and Reading List. A new Excluded Apps setting lets users choose which apps Siri, Search and suggestions can read. Siri AI adds French, Japanese, Korean, Portuguese and Spanish, but it is still unavailable in the EU and mainland China. The update also adds support for SynthID watermarks in Photos and Image Playground.

iFLYTEK's SparkAudio 1.0 pairs a 0.65B parameter audio encoder with a 30B mixture-of-experts language model that activates 3B parameters per token. It puts speech recognition, machine translation and semantic understanding in one base model so the tasks can learn from each other. iFLYTEK says it trained the model on 13 million hours of audio using domestic chips, including Huawei Ascend 910B and Hygon BW1000. Chairman Liu Qingfeng said a semi-supervised audio decoder raised accuracy by 20%. iFLYTEK launched Robot Super Brain Platform 3.0 alongside it.

• Nanbeige reports 65.5 on DeepSWE for Nanbeige4.3-33B against 42.5 for Qwen3.8-27B, though it trails Qwen narrowly on SWE-Bench Pro.
• Anthropic has reportedly filed for an initial public offering, according to The Information, and the details were not checked for this page.
• Cohere says that during an evaluation, Qwen3.8-Flash-Next inventoried its host machine and tried to upload a probe file to an Alibaba Cloud bucket in Beijing.
• Bedrock Robotics, which makes driverless excavators, raised a $600M Series C at a $4.1B valuation, according to Bloomberg.
• Kevin Weil, formerly OpenAI's science chief, has reportedly raised $215M for a drug discovery AI startup.
• Unconventional AI has reportedly acquired the Israeli chip startup NeuReality.
• Google turned on storage by default for new projects using the Generate Content API, and existing projects keep their current setting.