OpenAI and Anthropic clash in the 2026 AI mega IPO race despite lacking profit

OpenAI’s annual recurring revenue (ARR) of $25 billion as of March 2026 still trails Anthropic’s $19 billion by only $6 billion, a gap that the Korean‑language report from qjc.app notes could vanish by mid‑2027 given Anthropic’s 14‑fold year‑over‑year growth versus Open‑AI’s 3.5‑fold pace【report】. The disparity in valuation—$850 billion for OpenAI versus $380 billion for Anthropic—therefore reflects not just scale but a divergent capital‑efficiency trajectory. PitchBook’s comparative analysis ranks OpenAI’s “business quality” at the bottom of a three‑company set that includes Anthropic and Databricks, despite its premium market cap, underscoring investor concern that the higher price is not matched by operational performance【report】.

Cash burn further widens the strategic chasm. OpenAI is projected to lose $14 billion in 2026, with cumulative losses of $44 billion from 2023 through 2028 and a peak annual burn of $85 billion in 2028. HSBC estimates that OpenAI will need more than $207 billion in fresh capital to sustain operations through 2030, a “black‑hole” cash‑flow pattern that would make it the largest corporate burn in history【report】. By contrast, Anthropic’s 2025 EBITDA loss of $5.2 billion is modest, and its break‑even horizon is set between 2027 and 2029—at least a year earlier than OpenAI’s target. The report highlights Anthropic’s superior capital efficiency: $1 of capital raised generates $0.23 of ARR, double OpenAI’s $0.11, while OpenAI’s efficiency has deteriorated from $0.31 to $0.11 over the past 18 months【report】.

Revenue composition also diverges sharply. OpenAI’s income is still heavily consumer‑driven, with 75 % of its sales coming from ChatGPT subscriptions and the remaining 25 % from API usage. The platform boasts 900 million monthly active users, yet only 5.5 % convert to paying customers, meaning a substantial portion of its infrastructure supports free traffic. Moreover, its enterprise API market share has slipped from 50 % to 25 %, indicating erosion in the higher‑margin B2B segment【report】. Anthropic, by contrast, has built its model around enterprise contracts from the outset, limiting exposure to free‑user costs and aligning its cost structure with higher‑margin revenue streams. This focus is reflected in its ARR per dollar of capital raised, which remains stable at $0.23, whereas OpenAI’s metric has halved in the same period【report】.

The growth outlook is tempered by the looming profitability question. Both firms plan IPOs in Q4 2026, yet neither is profitable at the filing stage, as confirmed by Perplexity’s summary of internal financial documents【Perplexity】. Analysts cited in the qjc.app piece warn that the “mega‑IPO” narrative may be premature if cash‑burn trajectories are not curbed. OpenAI’s reliance on massive data‑center expansion to support its consumer base inflates operating expenses, while Anthropic’s tighter capital deployment suggests a more sustainable path to earnings. The disparity in cash‑burn efficiency, ARR growth, and revenue mix will likely shape investor sentiment when the two roadshows commence, with capital‑efficient enterprises such as Anthropic potentially commanding a premium despite a lower headline valuation.