OpenAI pushes Codex

OpenAI is no longer just selling Codex as a coding model; it is recruiting big consulting firms to wire the tool into large companies’ software operations. (reuters.com) Reuters reported on April 21 that OpenAI is expanding work with Accenture, Capgemini, CGI, Cognizant, Infosys, PwC and Tata Consultancy Services. OpenAI also said it is launching Codex Labs, which will place its own specialists inside customer organizations to connect Codex to existing systems and workflows. (reuters.com) OpenAI said Codex weekly usage rose to more than 4 million developers in recent weeks, up from about 3 million earlier in April. The company describes Codex as a tool that can write, review and reason about code across parts of the software development lifecycle. (reuters.com, (help.openai.com)) The pitch to large companies has shifted from model access to implementation. In February, OpenAI announced “Frontier Alliances” with Boston Consulting Group, McKinsey, Accenture and Capgemini, pairing consultants with OpenAI’s forward-deployed engineers to get enterprise customers to actually use its systems. (techcrunch.com) That puts consulting work — integration, process redesign and internal rollout — closer to the center of the AI sales cycle. TechCrunch reported in February that OpenAI was betting consultants could help customers change workflows and prove a return on investment, not just attach a model to existing tools. (techcrunch.com) OpenAI is making the same argument in pricing and product design. Its Codex rate card, updated April 2, moved many plans from per-message estimates to token-based pricing, and the help page says Codex usage averages about $100 to $200 per developer per month, with costs varying by model, automations and fast mode. (help.openai.com) Consulting partners are already presenting Codex as a workflow product, not just a chatbot for programmers. Cognizant said on April 21 that it is embedding Codex across its Software Engineering Group and using it for code generation, refactoring, testing, documentation and legacy-system modernization. (news.cognizant.com) Reuters framed the push against a more crowded market, with Anthropic gaining corporate traction for coding and reasoning and Microsoft, Google and Amazon investing heavily in business-focused artificial intelligence products. OpenAI’s answer is to put more of the sales effort into the hard part after the demo: getting a company’s developers, managers and internal systems to work with the tool every day. (reuters.com)