OpenAI and Anthropic offer millions in compute credits
OpenAI, Anthropic and cloud providers are giving startups token and cloud compute credits worth up to millions to win early customers.
TL;DR
- 01OpenAI, Anthropic and cloud providers are giving startups token and cloud compute credits worth up to millions to win early customers.
- 02OpenAI and Anthropic are offering startups free token and cloud compute credits worth millions, in a competition to lock early business customers into their ecosystems.
- 03Some individual startups have received competing offers that total more than $3 million, according to a Wall Street Journal report.
OpenAI and Anthropic are offering startups free token and cloud compute credits worth millions, in a competition to lock early business customers into their ecosystems. Some individual startups have received competing offers that total more than $3 million, according to a Wall Street Journal report.
How big are the offers and who is being targeted?
The offers range from hundreds of thousands to multiple millions, and Y Combinator cohorts are a primary target. A Google Cloud spokesperson said the company hands out up to $500,000 in credits, plus early access to Gemini models and occasional access to DeepMind engineers. Hans Ibarra, founder of AI voice startup Dialogus, said competing offers he received sometimes add up to more than $3 million. Sam Altman announced $2 million in token credits in exchange for equity stakes for Y Combinator companies. Anthropic countered with $500,000 and no equity, which the company said was a big jump from its previous $30,000 offer. OpenAI matched the $500,000 no-equity offer and added an optional $1.5 million in exchange for shares. The AI coding firm Cursor offered a 75 percent discount through July 5, illustrating that discounts and credits come in various forms.
Why are companies handing out so much free compute now?
The credit war coincides with margin pressure ahead of expected public listings for major AI firms. OpenAI and Anthropic, along with cloud providers, are offering steep discounts and large credits even as they face competition from lower-cost and open models, including many out of China. Anthropic saw revenue surge late last year thanks to Claude Code and Cowork, while OpenAI released GPT-5.4 in March and began actively selling its Codex tool to startups. The offers aim to make tools sticky and encourage startups to build on particular stacks before costs or switching friction rise.
How extensive could these giveaways become?
If the outreach continues at scale, the financial commitment could be substantial: with roughly 200 companies per Y Combinator cohort and four cohorts per year, the two firms could hand out up to $800 million in credits combined, based on the matched $500,000 offer figure. That projection comes directly from the observed deal sizes and YC cohort counts cited in the Wall Street Journal coverage.
Why it matters
This is a customer-acquisition tactic that shifts near-term economics. Startups receive immediate runway in compute and tokens, which can accelerate product development. For OpenAI, Anthropic and cloud partners, these credits trade near-term margin for customer lock-in that may raise lifetime value later. The tactic also raises the stakes in recruitment of the most promising early-stage AI firms, particularly those graduating from Y Combinator cohorts.
What to watch
Track whether the matched $500,000 no-equity offers persist across more YC cohorts and whether firms convert credits into paid usage after trial periods end. Watch public filings and margin disclosures from OpenAI and Anthropic as they pursue IPOs and need to show improving unit economics. Also monitor uptake by startups: if many accept and remain platform-locked, the strategy will have achieved its aim; if credits are used then startups switch providers, the approach will strain margins without securing long-term customers.
Source: Wall Street Journal
Written by The Brieftide · Source: The Decoder
The Brieftide Daily · 06:00
Briefs like this one, in your inbox every morning.
Continue reading
More in Open Source AIOpenAI joins Appia Foundation to build shared AI standards
OpenAI supports evaluation frameworks, safety practices and global cooperation through the Appia Foundation.
Zhipu AI GLM-5.2: 1M-token context, closes gap with Opus 4.8
GLM-5.2 ships under the MIT license with a stable one-million-token context and scores 74.4% on FrontierSWE, one point behind Opus 4.8.
OpenAI: PRC-linked influence operations target US AI debates
OpenAI says PRC-linked campaigns are using AI to push narratives on U.S. tech debates, data centers, tariffs and false ChatGPT claims.
OpenAI: LSEG scales trusted AI, empowers 4,000 staff
LSEG uses OpenAI to scale trusted AI across its global business, accelerating insights, shrinking release cycles and empowering 4.