What happened
OpenAI said it rewrote its Habitat storage platform from Python to Rust in Q2 2026 with 2 engineers, Codex, and GPT-5.5. The Rust service handles 95% of production requests and is 6x more CPU efficient.
Why it matters
Habitat now handles over 70 million requests per second for more than 1 billion weekly users across almost 40 regions, up from a single-database Python library two years ago.
What to watch
OpenAI says it will deprecate Python entirely in the coming weeks; the test is whether the Rust service keeps the same reliability. Watch the promised future post on multi-tenancy and Azure Cosmos DB scaling.
WHO IT HITSEngineering leaders at companies running large-scale online services get a concrete data point on rewriting performance-critical Python services in Rust, with OpenAI reporting 6x CPU and 15x memory efficiency gains. OpenAI's own product teams are the immediate beneficiaries, since the platform handles 95% of production requests.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
Habitat began in mid-2024 as a simple Python client-side library that interacted with ChatGPT's main server and Azure Cosmos DB. It saw rapid adoption among OpenAI product engineers, with no central push away from self-serve Postgres and Azure Cosmos DB. By the middle of 2025, it had reached its limits as a client-side implementation, as backward-compatible protocol changes had become infeasible and coordinating deployments across dozens of services took days.
OpenAI then pulled Habitat into its own service, accepting Python's performance trade-offs as a "strategic incursion of technical debt" to unblock product developers and achieve platform stability. The team made a calculated wager that rapid advancement of OpenAI's own coding models would simplify an eventual migration off Python, and OpenAI says that bet proved correct. In Q2 2026, with just 2 engineers, Codex, and GPT-5.5, the entire service was rewritten in Rust.
The outcome now hinges on whether the Rust service can maintain the reliability OpenAI says it needs while handling 95% of production requests, and whether the promised deprecation of Python proceeds as stated. OpenAI says it will share more learnings in a future blog, including on multi-tenancy reliability and scaling Azure Cosmos DB.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Dynatrace acquired Arize AI, adding AI observability, evaluation and agent monitoring to its application obser…
A Daily Dose of Data Science test kept LoRA adapters separate from a shared 7B base model, cutting 100 fine-tu…
A report by Spencer Kitts, Thomas Larsen and Sydney Von Arx says an OpenAI agent swarm very likely ran an atta…
Simon Willison wrote that many people, himself included, have gone through an existential crisis when a coding…
Stephen Aarons, a New Mexico defense lawyer of over 40 years, was held in direct contempt and fined $5,000 for…
Perplexity cofounder and Chief Strategy Officer Johnny Ho said GPT‑6 Astra can craft communications, edit real…
