Products·2 min read

Coding Freebie War: OpenAI Offers Two Months of Free Codex, Anthropic Lifts Claude Code Limits 50% the Same Day

Sam Altman pitches enterprises switching from Claude with 60 days of complimentary Codex access. Hours later, Anthropic boosts Claude Code weekly limits by 50% through July 13 — for free.

Coding Freebie War: OpenAI Offers Two Months of Free Codex, Anthropic Lifts Claude Code Limits 50% the Same Day
Share:

The AI coding tools war turned into a giveaway on May 13, 2026, when OpenAI and Anthropic unveiled back-to-back promotions clearly aimed at one another. OpenAI announced two months of free enterprise Codex access for companies that sign up within the next 30 days, while Anthropic, in a near-simultaneous post from its @ClaudeDevs account at 2:19 PM San Francisco time, raised Claude Code weekly usage limits by 50% through July 13, 2026, for every paid tier of its developer product. Business Insider dubbed the dual move a "freebie war."

OpenAI's offer is the more aggressive of the two on price. Sam Altman, posting on X, said: "Codex is the best AI coding product and we want to make it easy to try." The company is dropping seat fees for the Codex business tier during the trial and pitching the promo at companies disenchanted with Anthropic's recent decision to charge separately for third-party agent tools that piggyback on Claude Code subscriptions — a policy change that generated significant developer backlash on Hacker News and X. Enterprises that sign up before mid-June get the full two months free; with the 30-day window factored in, the campaign effectively keeps the door open into August.

Anthropic's response is not a price cut but a capacity expansion — though for power users it is arguably more valuable. The 50% increase applies to Pro ($20/month), Max 5x ($100/month), Max 20x ($200/month), Team ($30/user/month), and seat-based Enterprise plans. New caps roughly translate to about 60 weekly hours of Sonnet 4.6 plus 12 hours of Opus 4.7 on Pro, around 75 weekly hours of Opus 4.7 on Max 5x, and approximately 300 weekly hours of Opus 4.7 on Max 20x. The free Claude Code tier is excluded, and the boost activates automatically — no user action required. The promotion runs until July 13, exactly two months.

The competitive subtext is hard to miss. Anthropic has now made three consecutive capacity interventions in five weeks: the April 16 "Claude 2x" promo timed to the Opus 4.7 launch, a hourly limit doubling on May 6 (paired with a SpaceX compute deal), and now the weekly-limit jump. OpenAI, meanwhile, is betting that Codex's reportedly lower token consumption — some analyses peg it at 3-4x fewer tokens for equivalent tasks — combined with a zero-cost trial will pull workloads away from Claude. With Anthropic's revenue run rate hitting $30 billion and OpenAI eyeing $2.5 billion in ads revenue this year, both companies have the war chest to keep subsidizing developer trials. For coders, the next two months are an unusually good time to be picky.

More on Claude Code

Evergreen coverage we keep current — start here.

Want AI news before everyone else?

The morning's most important AI stories, straight to your inbox. No fluff.

Related Articles

OpenAI DevDay 2026: a new working world An original editorial illustration: a glossy black Dots agent bearing the OpenAI emblem stands on a mint pedestal. A looping track connects a cloud laptop running Codex, a golden GPT-6.1 Sol chip, and a violet Pro 500 Ultrafast module. Smaller agents and a completed task sheet orbit the central dot. OpenAI DevDay 2026 Codex dots GPT-6.1 Sol Pro 500 ULTRAFAST 29 SEPTEMBER BITSMINDS.COM
Products

OpenAI DevDay 2026: Dots, GPT-6.1 Sol and a $500 Plan

GPT-6.1 Sol: a brighter engine for a fifth of Astra's token price An imagined emerald and brass precision engine holds a golden solar core with the OpenAI knot. A fine orbital path leads toward a distant violet Astra star. Beside the engine, a machined price rail shows one bright gold token and four subdued green tokens, with the label 1/5 of Astra's standard API token price. GPT-6.1 Sol is engraved on the engine's pedestal. The orbital path is a conceptual illustration, not a quantitative benchmark; the article reports close performance on some tasks and remaining gaps on others. The ratio refers to standard input and output prices, not cached input, long-context rates or cost per task. Original vector illustration for BitsMinds, 30 September 2026. Source article: https://www.bitsminds.com/news/gpt-6-1-sol-launch-price-benchmarks . Static, self-contained SVG. ASTRA GPT-6.1 Sol 1/5 ASTRA'S STANDARD API TOKEN PRICE BITSMINDS.COM
Products

GPT-6.1 Sol: Near-Astra Scores for a Fifth of the Price

OPEN AGENT SAFETY PLATFORMBITSMINDS.COM
Products

Nvidia Moves the Agent Kill Switch Into Silicon