Thibault “Tibo” Sottiaux leads Product & Platform at OpenAI, covering ChatGPT, Codex, and the API platform. His work connects model development with the interfaces and engineering workflows that turn model capabilities into products. Before OpenAI, he worked at Google and Google DeepMind.[1][2]
Start with OpenAI’s Tibo Sottiaux: We really must get this right →
Education
Computer science, computational mathematics and applied mathematics
[1]- Also known as
- Tibo Sottiaux[1]
Links
X LinkedIn Masters of Scale biography Product & Platform lead · OpenAI
Background
Sottiaux’s publisher biography records degrees in computer science, computational mathematics, and applied mathematics from Université catholique de Louvain. Before OpenAI, he worked as a Google software engineer and later at Google DeepMind, where he led human-data work for Gemini and built infrastructure for research workflows. That experience concerned how researchers obtain useful training inputs and run their work, rather than only how an application looks.[1]
Career
Sottiaux led the team building Codex before taking responsibility for OpenAI’s broader product and platform organization. In an October 2026 interview, he describes a remit covering ChatGPT, specialist products such as Codex, and the APIs through which other developers use the company’s models. His Lenny’s Podcast appearance the same month identifies him as leading ChatGPT and Codex.[2][3]
Contributions
The 2025 Codex launch illustrates the type of product his team built: an agent receives a task, works in an environment containing the project, changes files, and runs checks before returning work for review. Its value comes from connecting model output with the surrounding software tools and evidence of execution. The launch article is now explicitly marked outdated, so it is useful as a historical design explanation rather than current setup instructions. Sottiaux is also named among the many contributors to the OpenAI o1 System Card, a report on the reasoning models’ safety evaluation.[5][2][6]
Ideas
In his October discussion with Lenny Rachitsky, Sottiaux argues that agents will increasingly take actions on the internet and discusses what that shift could mean for builders. It is a product leader’s forecast, not a measurement of present internet activity. The same conversation covers how coding and workflow tools might change as models improve. For newcomers, these interviews provide a way to examine the decisions behind an agent product: which tasks it takes on, how it fits into existing work, and where people still supply direction.[3]
Tools, writing & talks
Writing, research & talks
Start here
OpenAI’s Tibo Sottiaux: We really must get this rightmastersofscale.comAn October 2026 interview with Rana el Kaliouby about the scope of his product role and the relationship between research and products. It is a useful first stop for hearing his own account of what the team builds and why.
A recent, accessible conversation explaining his product remit and how model research becomes an agent people can use.
Interview guest[2]OpenAI's Codex team re-shipped Codex cloud, and Tailscale announced Codex Cloud can now securely connect to resources on a tailnet, opening private infrastructure to cloud coding agents.
Catalog workOriginal sourceOpenAI put its Decisions API into public beta for all developers. It is a realtime classifier for choosing the right model, tool, or action, with a claimed speedup of up to 10x over GPT-6 Luna via the Responses API.
Catalog workOriginal sourceOpenAI consolidated API usage tiers from five to three: Build, Launch, and Grow. The new top tier, Grow, requires $500 in total API payments, down from $1,000 for the prior highest tier, easing access to higher rate limits.
Catalog workOriginal source- OpenAI’s Head of ChatGPT: We’re entering a new era of AI (again) | Tibo Sottiauxlennysnewsletter.com
A conversation with Lenny Rachitsky about agent workflows and changing demands on builders. The public episode page supplies the topic guide and listening links; its written takeaway section is subscriber-only.
Interview guest[3]Original source OpenAI says GPT-6.1 Sol is its most demanded model ever. Capacity was added for ChatGPT and Codex, with speeds expected to improve to nearly twice the previous day's.
Catalog workOriginal sourceOpenAI relaunched Codex Cloud with configurable cloud environments and put the Agents API, which powers its cloud agents including dots and supports computer use, into preview.
Catalog workOriginal sourceOpenAI's Decisions API, powered by the Luna model, handles constrained decision making with visual inputs and is tuned to respond in a few hundred milliseconds end to end.
Catalog workOriginal sourceOpenAI introduced GPT-6.1 Sol, positioned near Astra in intelligence at one fifth the price with a 95% cache read discount. An ultrafast 8X speed option exists for Astra now and is coming to 6.1 Sol.
Catalog workOriginal sourceOpenAI announced ChatGPT Space, collaborative rich notes with live visualizations that people and agents edit together and that agents keep updated in the background, pitched as AGENTS.md taken further.
Catalog workOriginal sourceOpenAI's dots are 24/7 agents with their own computer and browser, powered by Astra and connected to over 4,000 apps. A primary dot is included in Pro without drawing usage, with multi-dot teams coming.
Catalog workOriginal sourceOpenAI's Codex lead explains that Pro $200 reopens with usage calculation netting out at half the prior API dollar value, no 5-hour limit, and GPT-6 Sol and Luna API prices cut 50%.
Catalog workOriginal sourceOpenAI released GPT-6 Sol and Luna with broad quality gains and a permanent 50% cut to API pricing, plus a one-time banked usage reset for Plus, Pro, and Business subscribers.
Catalog workOriginal sourceAn OpenAI employee lists several products shipping in the same week: Images 2.5, a live-voice model called GPT-Live-1, a new Agents API, a Data Agent, and ChatGPT for Financial Services, ahead of DevDay.
Catalog workOriginal sourceOpenAI is retiring GPT-5.3-Codex-Spark next week due to declining usage, pushing Codex users toward newer models.
Catalog workOriginal sourceOpenAI opened its Agents API, the same Codex-harness infrastructure behind ChatGPT Work, letting developers spin up scaled cloud agents with custom tools and sandboxes in under a minute.
Catalog workOriginal sourceOpenAI is pausing new sign-ups for its $200/month Pro plan, which grants access to GPT-6 Astra, due to system strain, while leaving other plans, the API, and existing Pro accounts untouched.
Catalog workOriginal sourceOpenAI's Tibo Sottiaux says paid ChatGPT subscribers accrue one banked usage reset for every day Astra is unavailable to them, counting from today, with the first credit landing in about three hours as the staggered rollout continues.
Catalog workOriginal sourceOpenAI has started releasing GPT-6 Astra in stages over several days while new serving infrastructure comes online. Plus users are included alongside Pro, Business and Enterprise. The post gives no capability or pricing details, and the linked release page could not be read.
Catalog workOriginal sourceOpenAI's Tibo Sottiaux spells out how Codex subscription multipliers are metered: the Pro tier's 20X is a weekly quota multiple of Plus, and neither Codex Pro plan enforces a rolling five-hour cap. Relevant context if you are comparing coding-agent plans on throughput.
Catalog workOriginal sourceOpenAI resets paid Codex and ChatGPT Work usage and enumerates the bugs that were quietly eating allowances: images kept through compaction, Stop-hook-trapped memory workers, /goal running past its stop condition, over-frequent automations, and subagents escalating to pricier models.
Catalog workOriginal sourceOpenAI is ending its partnership with Cursor and cutting off model access on November 12, citing trust. Own-API-key usage and OpenAI's IDE extensions remain supported paths, and the company says it will keep working with other harnesses.
Catalog workOriginal sourceOpenAI adds a team-oriented tier alongside the $100 Pro plan: full Codex and ChatGPT Work access, Workspace/Slack/GitHub/M365 connectors, SAML SSO and MFA, centralized billing, usage analytics and spend controls, and no five-hour limit.
Catalog workOriginal sourceOpenAI is reinstating the 5-hour rolling window on Plus accounts for ChatGPT Work and Codex, to smooth compute load and stop newer users burning a full week of quota in a single session. Pro $100 and Pro $200 subscriptions stay exempt for the upcoming months.
Catalog workOriginal sourceOpenAI traces faster Codex quota drain to specific causes: degraded prompt-cache hit rates, inefficiency with images in long sessions that compact repeatedly, high p95 Computer History usage, and a conversation-title feature burning more than intended. Fixes shipped alongside a full reset for paid p
Catalog workOriginal sourceOpenAI's Codex lead says accounts reporting faster limit drain were largely running sub2api, converting subscription auth into resold API traffic, which trips fraud prevention. Signing in with ChatGPT from official or OSS clients such as Pi and OpenCode stays supported. Drain investigation ongoing.
Catalog workOriginal sourceTransparent-background image generation is now available for GPT-Image-2 in API preview and, according to an OpenAI engineer, in ChatGPT as well, making generated icons, stickers, and product assets directly compositable.
Catalog workOriginal sourceOpenAI previews Private Safety Processing: for Zero Data Retention customers, safety pattern-matching runs on infrastructure the customer controls and returns only limited signals, leaving prompts and responses unexposed to OpenAI employees. A hosted option with customer-controlled keys starts rolli
Catalog workOriginal sourceA Codex postmortem on destructive deletes: cleanup routines reused variables like $HOME and could target the real home directory, and temp paths were overwritten without checking contents. Fixes span explicit model instructions, execution-layer detection of high-risk delete commands, and making Full
Catalog workOriginal sourceAn OpenAI engineer documents the exact Codex config keys for running GPT-5.6 Sol at a one-million-token window, plus a single-session CLI override, with a warning that the shipped default was tuned deliberately for cost and performance.
Catalog workOriginal sourceAn argument that a token is not a standard unit: identical text costs different token counts on different tokenizers, so a cheaper per-token price can still produce a larger bill. One sample measured 766 tokens on GPT-5.6 Sol against roughly 1,170 for Claude Opus 5. Vendor-authored, so verify on you
Catalog workOriginal sourceOpenAI reset usage limits for all Codex and ChatGPT Work users after a near-global outage in the early hours of the morning.
Catalog workOriginal source- Introducing Codexopenai.com
The May 2025 launch explanation of a coding agent that edits and tests a repository inside an isolated environment. Read it for the original workflow and review model; follow the page’s current-product links for today’s instructions.
- OpenAI o1 System Cardarxiv.org
A collective report describing safety evaluations and external testing of OpenAI’s reasoning models. Sottiaux appears on its contributor list; the report does not identify a unique individual research contribution for him.
Named co-author in a large team[6]
Tools
- Codex Securitydevelopers.openai.com
CLI and TypeScript SDK from OpenAI that scans codebases for security vulnerabilities using Codex models.
Catalog workOriginal source