Why Anthropic and OpenAI Both Shouted '2x' at the Same Time
Anthropic's 'off-peak 2x usage' and OpenAI's '2x Codex rate limits' are running simultaneously from March 13-27. Both companies offering '2x' promotions at the same time is no coincidence. We analyze the fierce market share battle in the coding agent space and the landscape of 7 major players.
The battle for coding agent market share is heating up. Anthropic and OpenAI are running '2x usage' promotions at virtually the same time. Anthropic offers double usage during off-peak hours from March 13 to 27, while OpenAI provides doubled Codex rate limits from February 2 to April 2.
The two-week window from March 13 to 27 is when both promotions overlap precisely. Both companies putting forward the same '2x' number simultaneously is no coincidence. It's a reflection of the fierce competition for market share in the coding agent space.
Spare GPUs? Use More, Says Anthropic's Happy Hour
Anthropic's promotion runs for two weeks from March 13 to 27. The key is doubling the 5-hour usage limit during 'off-peak hours' only. Off-peak hours are defined as everything outside 8AM-2PM US Eastern time. For users outside the US, this effectively means doubled benefits throughout most working hours.
The scope is notably broad. It applies to Claude web, desktop app, mobile, Claude Code, Excel, and PowerPoint, covering virtually all Claude products. No signup or application is needed. It's automatic across all plans: Free, Pro, Max, and Team.
Anthropic has likened this promotion to a 'restaurant happy hour,' a strategy to redistribute peak-hour demand to off-peak hours. Alongside this promotion, they also announced the general availability of the 1 million token context window. Both Opus 4.6 and Sonnet 4.6 can now use 1M tokens at no additional cost. With usage doubled and context quintupled (from 200K), it's a significant value proposition for users.
OpenAI Fires Back With Two Months of 24/7 Full Throttle
OpenAI's promotion is more aggressive in both scale and duration. Running for a full two months from February 2 to April 2, it has no time restrictions like off-peak hours. The 5-hour rate limit and 7-day maximum usage for Codex are both doubled around the clock. OpenAI maintainer etraut-openai confirmed on GitHub that both the 5-hour and 7-day windows are doubled.
However, the scope is limited to Codex only. ChatGPT and other OpenAI products are not included. Paid plan users (Plus, Pro, Business, Enterprise, Edu) are the main targets, while Free and Go users receive one month of free Codex access.
OpenAI has been rapidly expanding the Codex ecosystem alongside the promotion. On February 13, they introduced a credit system allowing users to purchase credits and continue after hitting limits. March 4 saw the Windows native app launch, clearing a waitlist of over 500,000. On March 5, GPT-5.4 was integrated, bringing native Computer Use, 1M token context, 33% fewer errors, and 47% less token consumption.
Notably, Codex Security launched on March 6 and in its research preview phase has already scanned 1.2 million commits, discovering 792 critical and 10,561 high-risk vulnerabilities. On March 7, the Codex for Open Source program was announced, offering 6 months of free ChatGPT Pro to maintainers of projects with 1,000+ stars.
Same '2x' Label, Wildly Different Playbooks
On the surface, both promotions look similar, but their strategic intent differs. Anthropic is running a 'universal promotion' across Claude's entire ecosystem to reshape user habits, while OpenAI is going all-in on the coding agent vertical with a 'focused offensive.' Anthropic's approach is shorter but broader; OpenAI's is narrower but longer with no time restrictions.
Some users have questioned the effectiveness of OpenAI's promotion. GitHub issue #11785 includes complaints about the 2x benefit not being consistently applied.
Seven Contenders, From Claude Code to Kiro, All In
As of March 2026, seven major players are competing for market share in the coding agent space. Claude Code currently leads. According to Anthropic's Agentic Coding Trends Report, 63% of developers use Claude Code, an explosive jump from 4% in May 2025. It leads in satisfaction (46%) and VS Code extension installations at 5.2 million.
OpenAI's Codex, powered by GPT-5.4, reached 60% of Cursor's usage immediately after launch, showing rapid growth. Its VS Code extension has also hit 4.9 million installations.
Cursor differentiates with Background Agents and Automations features, along with JetBrains ACP integration. Windsurf was acquired by Cognition (Devin's developer) for $250M, reaching a $10.2B valuation with unique features like Arena Mode.
GitHub Copilot offers Agent Mode and an autonomous Coding Agent for PR creation, with a competitive $10/month price point. Google's Antigravity positions itself as an agent-first IDE based on Gemini 3 Pro, currently in free public preview. AWS's Kiro takes a spec-driven development approach with native Bedrock integration.
A $52 Billion Pie, and Everyone Wants the Biggest Slice
The coding agent market is growing explosively. According to market research, the coding agent segment is projected to grow from $7.84B in 2025 to $52.62B by 2030, at a CAGR of 46.3%. The broader agentic AI market is expected to grow from $9.14B in 2026 to $139B by 2034, at a CAGR of 40.5%.
Developer adoption is surging. Currently, 95% of developers use AI weekly, 75% use it for more than half their work, and 55% regularly use AI agents. Gartner predicts that 60% of new code will be AI-generated by the end of 2026.
In this market environment, the '2x promotions' are far more than simple discount events. Coding agents are tools that become more deeply integrated into a user's workflow the more they're used. Developers who get comfortable with a particular agent rarely switch. Both Anthropic and OpenAI understand this 'lock-in' effect well, which is why aggressive promotions are being deployed under the logic that securing early usage directly translates to market share.
