OpenAI and SpaceXAI dominated the news today. Sam Altman’s GPT-5.6, which is regarded as a model as capable as Anthropic’s Opus 4.7 and Fable 5, is finally out in the wild and available across three tiers with varying prices. They even released GPT-Live as an AI voice assistant, which Altman called “magical and real.” Elon Musk, on the other hand, went with a different approach for SpaceXAI’s Grok 4.5 model – the team designed it specifically for coding purposes, backed by the recently acquired Cursor team.
Just as the US assumed that it had taken a huge leap in the AI space, China’s ByteDance announced Seedream 5 Pro as an advanced imaging model whose performance is comparable to GPT Image 2 but is four times cheaper to use! Google, however, takes it upon itself to answer the foreign and homegrown competition by working hard to release its Gemini 3.5 Pro model.
Additionally, we also saw Apple starting to accept debit and credit cards in India for making payments for App Store and iCloud purchases. India’s TRAI also saw itself getting involved with the MeitY over displaying ‘spam’ labels on 140 and 160 series numbers.
For anyone following technology, it is an incredibly busy day. Hence, for all those who missed the updates, here is a simple, comprehensive breakdown of today’s top updates from technology and AI.
OpenAI releases GPT 5.6 Sol and GPT-Live
Sam Altman and OpenAI released the GPT-5.6 family officially, highlighting its flagship GPT-5.6 Sol model. Early testers and developers have noted that while Sol may not necessarily be “smarter” than Anthropic’s Fable, it is significantly faster, highly reliable, and produces exceptional overall work, leading teams to heavily increase their token usage. Alongside the rollout of GPT-5.6, OpenAI also introduced GPT-Live – a new generation of voice models designed for natural, real-time human-AI interaction. GPT-Live is actively rolling out in ChatGPT.
Grok 4.5 debuts via Cursor and SpaceX
In Elon Musk’s camp, Cursor and SpaceX have launched Grok 4.5. Musk and his team boldly claim that it delivers Opus 4.7 quality at a fraction of the cost. Processing at a blistering speed of 80 tokens per second, this model aims to democratise high-end compute access. The aggressive pricing structure is set at just $2 per million input tokens and $6 per million output tokens, competing with open-weight models.
ByteDance Seedream 5 Pro goes after visual processing
ByteDance also officially unveiled Seedream 5 Pro, its new image generation model that comfortably secures the global number two spot after OpenAI’s model. Coming close to the gold standard of GPT Image 2, Seedream 5 delivers its magic at a four times cheaper rate, costing between $0.045 and $0.09 per image. Specialising heavily in complex edits and crisp infographics, this newly released tool easily outperforms Meta’s Muse Image on benchmarks, establishing ByteDance as a massive contender in the creative AI space for global designers.
TRAI intervenes on Truecaller spam labelling
Moving from AI to regulatory news, the Telecom Regulatory Authority of India (TRAI) has issued an important directive to caller identification applications like Truecaller. The agency ordered these platforms to immediately cease displaying spam labels on the 140 series, reserved for telemarketing, and the 1600 series, utilised for vital banking communications. TRAI emphasised that these are official business numbers, noting that aggressive user reporting was inadvertently blocking genuine, critical calls that citizens actually need to receive on their phones every single day.
Apple restores card payments for Indian App Store
Apple has officially restored credit and debit card payment methods for App Store and iCloud purchases in India. After a long four-year hiatus, Apple followed strict updates to the Reserve Bank of India regulations back in 2022, consumers were forced to exclusively use UPI or net banking options. This newly implemented phased rollout now allows eligible Visa and Mastercard users to successfully link their accounts to Apple.
Google readying Gemini 3.5 Pro
Google is gearing up for Gemini 3.5 Pro, scheduled to debut on July 17. The tech giant has noticeably fallen behind the frontier pack (Anthropic, OpenAI, SpaceXAI) recently, creating intense pressure for this launch. Add to that a string of high-profile talent departures, and the industry is eagerly watching to see if Google can reclaim its innovation crown.
Prediction markets spike over rumoured GPT-6 release
Meanwhile, rumours regarding OpenAI’s next-generation powerhouse, GPT-6, are reaching a fever pitch across the tech community. Speculation spiked dramatically on prediction platforms like Polymarket, pointing toward an August 14 arrival date. Rumoured to be built upon a massive, completely redesigned retrain architecture, GPT-6 is explicitly taking direct aim at unseating the popular Fable model. If these rumours hold true, the AI landscape could experience another major shift before the end of the current summer season in August.