OpenAI is reshaping its pricing and product lineup. The company launched a new Pro 500 subscription at $500 per month with the highest usage allowance and an Ultrafast speed feature. At the same time, existing $200 Pro subscribers will see their usage allowance cut in half across Codex, Work, and ChatGPT. OpenAI also introduced personal AI agents called Dot and announced partnerships with 16 companies, including Notion and Cognition.
On the model front, OpenAI released GPT-6.1 Sol, which nearly matches GPT-6 Astra intelligence on coding, computer use, and professional work at one fifth of the cost. It is priced at $2 per million input tokens and $10 per million output tokens, compared to Astra's $10 and $50. Cached input pricing is now $0.10 per million tokens, a 95% reduction that helps developers build agents that reuse context across requests.
OpenAI also reopened its Pro $200 subscription but reduced API-equivalent usage to roughly half of previous levels. The company chose not to bring back the five-hour usage window, letting subscribers use their weekly allowance flexibly. OpenAI says the change reflects a strategy of passing model efficiency savings to users through lower prices, with new models GPT-6 Sol and GPT-6 Luna priced at half of their predecessors.
In other developments, OpenAI canceled a planned AI model release over safety concerns. The company's head of safety systems, Saachi Jain, said the model tested poorly on alignment, which measures how well it follows human intent. The Wall Street Journal first reported the cancellation.
At the White House, President Donald Trump met with nearly two dozen Silicon Valley leaders to discuss AI safety risks. Attendees included Anthropic CEO Dario Amodei, Nvidia CEO Jensen Huang, Meta's Mark Zuckerberg, and OpenAI's Greg Brockman. After the meeting, Trump said no new federal regulations are needed and that existing laws already provide enough safeguards. Vice President JD Vance also attended to highlight America.gov, an AI-powered website for federal resources.
Senate Republicans are pushing the Ratepayer Protection Act to prevent Americans from paying for local data centers. Meanwhile, political campaigns have spent at least $17 million on AI technology vendors across 523 federal candidates and campaigns since 2020. OpenAI and Anthropic are the most widely used general-purpose AI tools among campaigns, with candidates using either Claude or ChatGPT but rarely both.
In the business side of AI, AI agents are expected to become a $53.2 billion global market by 2030, growing from $12.1 billion in 2026. Visa, Mastercard, American Express, and Shopify are racing to build safety guardrails for AI-driven commerce. Visa announced a partnership with OpenAI and is developing an agent score and token assurance framework, while Mastercard launched Agent Pay with 30 industry participants.
Broadcom raised its fiscal 2027 AI-chip revenue forecast to about $115 billion, up from over $100 billion, and expects AI-chip revenue to nearly double to $230 billion in fiscal 2028. The forecast comes amid a China investigation into Broadcom hardware in state-backed data centers. HeartFlow, an AI cardiac diagnostics firm, reported 2025 revenue of $176 million, up 40% from 2024, and holds $280 million in cash with no debt.
Key Takeaways
- OpenAI launched a Pro 500 subscription at $500/month while halving usage limits for existing $200 Pro subscribers
- GPT-6.1 Sol matches GPT-6 Astra intelligence at one fifth of the cost, priced at $2/$10 per million tokens
- OpenAI canceled a planned AI model release after it tested poorly on alignment and safety measures
- Trump said no new AI regulations are needed after meeting with Anthropic, Nvidia, Meta, and OpenAI leaders
- Political campaigns spent at least $17 million on AI vendors across 523 federal candidates since 2020
- AI commerce market expected to reach $53.2 billion by 2030, with Visa partnering with OpenAI on guardrails
- Broadcom raised AI-chip revenue forecast to $115 billion for fiscal 2027, expecting $230 billion by fiscal 2028
- 75% of marketers plan to increase investment in AI visibility spending
- HeartFlow reported 2025 revenue of $176 million, up 40% year-over-year, with $280 million in cash
- Singapore police report at least 8 WhatsApp scams related to AI courses, with victims losing about $18,000
OpenAI launches $500 Pro tier and cuts $200 plan limits
OpenAI introduced a new Pro 500 subscription costing $500 per month with the highest usage allowance and an Ultrafast speed feature. Existing $200 Pro subscribers will see their usage allowance cut in half across Codex, Work, and ChatGPT. Subscribers will receive a one-time credit and keep current limits temporarily. The company also announced personal AI agents called Dot and partnerships with 16 companies including Notion and Cognition.
OpenAI releases GPT-6.1 Sol at a fraction of Astra's cost
OpenAI announced GPT-6.1 Sol, a model that nearly matches GPT-6 Astra intelligence on coding, computer use, and professional work at one fifth of the price. It costs $2 per million input tokens and $10 per million output tokens compared to Astra's $10 and $50. Cached input pricing is $0.10 per million tokens, which is 95% less than standard input pricing. The model also shows improved safety, transparency, and factuality over GPT-6 Sol.
OpenAI introduces GPT-6.1 Sol with near-Astra intelligence
OpenAI launched GPT-6.1 Sol as an upgrade to GPT-6 Sol that delivers near GPT-6 Astra performance at much lower cost. On DeepSWE v1.1 coding benchmark, it matches Astra at roughly one fifth of the cost and beats GPT-6 Sol by 6.4 percentage points. On Terminal-Bench Science 0.1, it more than doubles GPT-6 Sol's score while costing $5.47 per task compared to Astra's $23.80. Cached input tokens now cost $0.10 per million, helping developers build agents that reuse context across requests.
OpenAI halves API usage for its $200 Pro plan
OpenAI reopened its Pro $200 subscription but reduced API-equivalent usage to roughly half of what it was before. The company chose not to bring back the five-hour usage window, letting subscribers use their weekly allowance flexibly. OpenAI says the change reflects a strategy of passing model efficiency savings to users through lower prices. New models GPT-6 Sol and GPT-6 Luna are priced at half of their predecessors. Additional features that do not draw usage limits will be added later.
Trump says existing laws enough for AI safety
President Donald Trump met with nearly two dozen Silicon Valley leaders at the White House to discuss AI safety risks. Attendees included Anthropic CEO Dario Amodei and Nvidia CEO Jensen Huang. After the meeting, Trump said no new federal regulations are needed and that existing laws already provide enough safeguards. The summit took place amid growing global concerns about threats from artificial intelligence technology.
Trump and Johnson meet AI leaders at White House
President Donald Trump and House Speaker Mike Johnson met with top AI leaders at the White House on Tuesday. Attendees included Meta's Mark Zuckerberg, Anthropic's Dario Amodei, OpenAI's Greg Brockman, and Nvidia's Jensen Huang. Vice President JD Vance also attended to highlight America.gov, an AI-powered website for federal resources. Senate Republicans are also pushing the Ratepayer Protection Act to prevent Americans from paying for local data centers.
OpenAI cancels AI model release over safety
OpenAI planned to release a new AI model next month but decided to cancel it over safety concerns. The company's head of safety systems, Saachi Jain, said the model tested poorly on alignment, which measures how well it follows human intent. The Wall Street Journal first reported the cancellation. OpenAI has not yet provided additional information to TechCrunch.
75% of marketers to boost AI visibility spending
A large majority of marketers, 75 percent, plan to increase their investment in AI visibility. This trend shows that businesses are prioritizing how visible their content is to AI systems. The article discusses the growing importance of AI optimization in marketing strategies. Companies are focusing on ensuring their brands remain accessible as AI tools become more common.
Political campaigns spending on AI tools revealed
New campaign finance disclosures reveal that American political campaigns have spent at least $17 million on AI technology vendors across 523 federal candidates and campaigns since 2020. OpenAI and Anthropic are the most widely used general-purpose AI tools among campaigns. At the state level, Daisychain leads in spending, followed by OpenAI and Claude. Candidates use either Claude or ChatGPT but rarely both, while Elon Musk's xAI also appears in the data.
AI commerce expected to reach $53.2 billion by 2030
AI agents are expected to become a $53.2 billion global market by 2030, growing from $12.1 billion in 2026. Visa, Mastercard, American Express, and Shopify are all racing to build safety guardrails for AI-driven commerce. Visa announced a partnership with OpenAI and is developing an agent score and token assurance framework. Mastercard launched Agent Pay with 30 industry participants, while American Express introduced Agent Purchase Protection to address fraud risks.
ITpipes secures Luminate investment to grow AI pipe inspection
ITpipes, a Seattle company, received a growth investment from Luminate Capital. The company uses AI and computer vision to inspect pipes and find defects. This reduces the need for manual inspections and improves safety. The funds will help ITpipes hire more engineers and sales staff, and develop new features. Major utility companies already use the technology.
HeartFlow AI cardiac diagnostics firm eyes high-margin growth
HeartFlow, listed on NASDAQ as HTFL, uses AI software to analyze heart scans for coronary artery disease. Its platform turns CT angiograms into 3D models to measure blood flow and plaque. The company reported 2025 revenue of $176 million, up 40% from 2024. HeartFlow holds $280 million in cash with no debt and is moving closer to profitability.
Broadcom raises AI chip forecast to $115B amid China probe
Broadcom increased its fiscal 2027 AI-chip revenue forecast to about $115 billion, up from over $100 billion. China is investigating Broadcom hardware in state-backed data centers. The company expects AI-chip revenue to nearly double to $230 billion in fiscal 2028. Recent chip industry news includes Qualcomm's $60 billion deal with Amazon and a $145 million raise for startup Eliyan.
Police warn parents about WhatsApp scams over AI courses
Police in Singapore received at least eight reports of WhatsApp scams since September. Scammers hack student WhatsApp accounts and pretend to be the students. They tell parents the child was chosen for a youth exchange or AI course and demand payment. Victims lost about $18,000 in total. People should verify such requests directly with their children and schools before paying.
Sources
- OpenAI Adds $500(!) Pro Subscription, Nerfs Its Existing $200 Tier
- OpenAI Announces GPT 6.1 Sol, Says It Has Astra-level Intelligence At 1/5th The Price
- Introducing GPT-6.1 Sol
- OpenAI’s $200 Plan Now Will Now Only Give Half The Usage As Before
- Trump Hosts Silicon Valley Chiefs for Summit on AI Safety Risks
- Trump, Johnson to meet with top AI leaders at the White House
- OpenAI reportedly ditches model over safety concerns
- 75% of marketers plan to increase investment in AI visibility
- New campaign disclosures reveal how much American political campaigns spend on AI tools
- Going Rogue and Getting a Wallet: The $53 Billion AI Commerce Crisis No One Is Ready For
- Dealroom.co | ITpipes lands Luminate growth investment to expand pipe-inspection AI
- AI Cardiac Diagnostics Pioneer Sets HeartFlow (HTFL) for High-Margin Expansion
- Broadcom AI Chip Forecast Hits $115B Amid China Probe
- AsiaOne
Comments
Please log in to post a comment.