OpenAI On Monday, it began emailing more than 8,000 developers who applied for it invitation only GPT-5.5 side with a surprise consolation prize: a tenfold increase in Codex rate limits on personal ChatGPT accounts, effective immediately and lasting until June 5th.
"Over 8,000 people inquired in just 24 hours, and we wish our office was big enough to accommodate everyone, but we couldn’t accommodate every applicant." the company wrote in an email obtained by VentureBeat. "As a small token of appreciation, we’ve increased your Codex rate limits by 10x on your personal ChatGPT account until June 5th."
The gift is not limited to the lucky few who get invited to the party. Everyone who raised their hand — whether they were rejected, waitlisted or turned away — received a rate limit increase, according to the email, which was confirmed by multiple buyers on social media.
CEO Sam Altman telegraphed the move at X shortly before inboxes started lighting up. "We’ll do something nice for anyone who applies for a GPT-5.5 party and we don’t have room." he wrote. "We hope you enjoy!" The post garnered more than 521,000 views within hours.
What a month of extra loaded Codex access actually means for developers
The practical implications are huge. CodexOpenAI’s AI-powered coding agent operates under daily usage limits that vary by subscription level. A tenfold increase in these caps gives developers more room to prototype, debug, and ship code. GPT-5.5 — Matches the per-token latency of GPT-5.4 while performing with higher levels of intelligence and using significantly fewer tokens to perform tasks, according to OpenAI.
A 31-day window is generous enough to reshape habits. By empowering thousands of developers with expanded access during a critical adoption period, OpenAI effectively subsidizes the deep, sustained use that turns exciting experiments into daily addictions. Once developers experience Codex to its full potential, they won’t want to go back – and when the limits reset on June 5th, a significant number will upgrade their subscriptions to protect the workflow they’ve built.
The developer community responded with joy and regret. "I literally don’t take my Codex hat off for a month," a developer He announced in X. Others kicked themselves for not signing up. "Not checking in last time since I’m not in SF," one wrote.
Several users have raised a question that OpenAI has yet to publicly answer: does the boost stack with the 20x multiplier of the existing Pro $200 tier? One user reported that OpenAI support says no – users get whichever limit is higher, not the combined total. "The key question is not whether the 10-fold increase is only for party candidates," they wrote "Depends on if it’s bundled with Pro or not."
OpenAI did not immediately respond to a request for comment on whether boosts stack with Pro-level limits.
Inside a low-key meeting an AI has planned for himself
The rate limit giveaway is a sidecar to the main event: "GPT-5.5 on 5/5," an invite-only gathering that continues tonight from 5:55 PM to 8:55 PM PDT at an undisclosed San Francisco location. OpenAI gave the evening’s account "A low-level meeting with Sam and the team behind GPT-5.5," promising food, drink, community, gifts and fun – not a product ad. Even the address was kept secret until invitations were confirmed—a touch of exclusivity that generated its own buzz.
In a detail that doubled as a product demo, Altman revealed that GPT-5.5 he planned the party himself. The model proposed a date of May 5, suggested that human developers give the toast, not AI, and recommended setting up a suggestion box for the next-generation model. This is how Altman described it "strange behavior." With the code governing the selection process, registrations closed shortly after opening due to overwhelming demand.
Altman also sent an unexpected invitation. He publicly asked Elon Musk to participate, saying: "It can come if it wants to… the world needs more love.” The gesture comes amid Musk’s ongoing lawsuit against OpenAI seeking up to $150 billion in damages — making the invitation look less like diplomacy and more like performance art.
Anthropic’s competitive reception makes planning overlap a Silicon Valley spectacle
This is where the story gets interesting. Anthropic is hosting its own invite-only event in San Francisco on Tuesday evening, VentureBeat has confirmed. "Media VIP welcome reception" Around the same times as the OpenAI batch. The reception serves as a warm-up for Anthropic’s Code with Claude developer conferencethe company’s second annual meeting focused on its API, CLI tools, and Model Context Protocol (MCP). The corresponding conference will take place tomorrow.
It’s hard to dismiss the scheduling overlap as a coincidence. Both companies are holding developer-focused events on the same evening, in the same city, targeting many of the same people. Whether it’s deliberate counterprogramming or genuine coincidence, the optics neatly capture what stands in the way of the industry’s most consistent competition.
At Anthropic’s conference, its executive and product teams will discuss Claude Code, agent implementation strategies and product roadmap – all aimed at the same newly adopted developer audience. One month of free Codex upgrades from OpenAI.
How Anthropic beat OpenAI in terms of revenue – and what it means for the coding wars
The dueling cocktail hours are the social manifestation of a more consequential battle over revenue, developer acceptance, and investor confidence—one that is heavily tilted in Anthropic’s favor.
according to Counterpoint Research dataAnthropic surpassed OpenAI for the first time in global LLM revenue market share in Q1 2026, with 31.4% compared to OpenAI’s 29%. But the nearby hood hides the dramatic structural difference. According to Counterpoint estimates, Anthropic achieved this share with about 134 million monthly active users, compared to about 900 million for OpenAI — $16.20 per active user for Anthropic and $2.20 for OpenAI. OpenAI provides large-scale commands; Anthropic generates about seven times more revenue per user. That gap is the central tension in this competition.
The enterprise queue has been established for more than a year. Menlo Ventures — whose portfolio includes Anthropic — estimates the company now accounts for 40% of LLM spending, down from 24% last year and 12% in 2023, while OpenAI’s share has fallen to 27% from 50% in the same period. Anthropic has had an almost unprecedented 18 months on the coding LLM leaderboard, starting with Claude Sonnet 3.5 in June 2024. This edge in code – AI’s first true killer software – has been the engine behind wider enterprise adoption and Anthropic’s revenue growth.
The top numbers tell the rest of the story. Anthropic said earlier this month annual revenue exceeded 30 billion dollarsMore than $9 billion by the end of 2025, with more than 1,000 business customers now spending more than $1 million a year — a figure the company said has more than doubled since February.
That’s the run rate, sources familiar with Anthropic’s financial statements told TechCrunch currently it is close to 40 billion dollarsdriven mainly by demand for Claude Code and Cowork. OpenAI, meanwhile, came out on top Annual income of 25 billion dollars As of February, according to Reuters – but the Wall Street Journal reported that the company recently missed its forecasts for user growth and revenueCFO Sarah Friar warned colleagues that unless growth accelerates, the company may struggle to fund future computing contracts.
It has moved to finance at a pace that could redraw the industry’s power map. Anthropic raised $30 billion in February at a $380 billion valuation. Bloomberg reported last week that the company has begun weighing a new funding round that will value it More than 900 billion dollarspotentially leapfrogging OpenAI as the world’s most valuable AI startup. OpenAI was valued at $852 billion at the end of March after breaking records $122 billion funding round. If Anthropic were to move forward on the terms described, the company would not only more than double its value, but also surpass OpenAI — a turnaround that would have seemed unthinkable six months ago.
Two parties, two visions and one city at the heart of the AI industry’s defining competition
For over 8,000 developers who applied GPT-5.5 sidethe immediate value is simple: a full month of fully expanded Codex usage, for free, at a time when both companies are shipping at high speeds. It’s harder for the industry to miss the signal. Two of the world’s most valuable private companies compete for developer loyalty with a combination of free bonuses, invite-only parties, celebrity CEO engagement and multibillion-dollar enterprise initiatives — all in the same 24-hour window, in the same seven-square-mile city.
Wider stakes go well beyond cocktail napkins and fare limits. Both companies are moving toward a potential IPO. Both are seeking the same Wall Street backers for joint ventures. Both are competing to define how and by whom the next generation of software will be built. The rest of the developers are the beneficiaries of a spending war that shows no sign of cooling off for now.
Tonight in San Francisco, the Anthropic reception starts at 5:00 p.m. The OpenAI party starts at 5:55 p.m. VentureBeat will be at both. And somewhere between the two spaces, the 8,000 developers who can’t get into either room will burn through the new rate limits — building the future with the model they first opened.
Michael Nunez is a VentureBeat editor covering artificial intelligence. He’s attending both the Anthropic Codex VIP welcome reception with Claude Media and the OpenAI GPT-5.5 launch event in San Francisco tonight.
This story is evolving and will be updated.
