On Wednesday, Sam Altman-led OpenAI launched Codex CLI, a new open-source coding agent, along with two AI reasoning models.
What Happened: The newly launched open-source coding agent is OpenAI's most notable open-source release since 2019.
The AI startup also unveiled the full-scale version of its "o3" model--described as its most advanced AI system to date--and a more compact, efficient variant known as "o4-mini."
Taking to X, formerly Twitter, Altman said, "o3 and o4-mini are super good at coding, so we are releasing a new product, Codex CLI, to make them easier to use. This is a coding agent that runs on your computer. It is fully open source and available today; we expect it to rapidly improve."
See Also: Arm Aims for 50% Data Center CPU Market Share by 2025, Challenging Intel and AMD
Codex CLI is built to operate directly on a user's machine while connecting to the cloud to leverage the reasoning capabilities of its o3 and o4-mini models. Beyond generating code snippets, it can independently determine which tools to use locally to carry out tasks.
The company also noted that Codex CLI will soon support features from its recently released GPT-4.1 model.
To drive adoption, OpenAI has launched a $1 million grant program, offering selected developers $25,000 in API credits to build innovative projects using Codex CLI.
Moreover, according to OpenAI, training the o3 model required roughly ten times more computational resources than its earlier o1 model, previously considered its top performer in reasoning.
Subscribe to the Benzinga Tech Trends newsletter to get all the latest tech developments delivered to your inbox.
Why It's Important: This launch comes at a time when OpenAI faces mounting pressure from competitors such as China's DeepSeek, which in January launched its R1 model, an open-source AI, reportedly capable of mimicking OpenAI's advanced reasoning abilities.
In February, Anthropic also introduced a model capable of delivering both rapid, instinctive responses and more deliberate, step-by-step reasoning when necessary--a level of adaptive flexibility that OpenAI's models have yet to fully replicate.
Last month, Alphabet Inc.'s (NASDAQ:GOOG) (NASDAQ:GOOGL) Google released its Gemini 2.5 Pro, a reasoning-focused model that outperformed OpenAI's o3-mini across several benchmark tests.
OpenAI has also completed a historic $40 billion funding round, raising its valuation to $300 billion. Major contributors to the round include SoftBank Group (OTC:SFTBF) (OTC:SFTBY) and Microsoft Corp. (NASDAQ:MSFT).
This reportedly marks the largest funding round ever recorded and nearly doubles OpenAI's earlier valuation of $157 billion.
Photo Courtesy: Meir Chaimowitz on Shutterstock.com
Check out more of Benzinga’s Consumer Tech coverage by following this link.
Read Next:
- OpenAI And Microsoft Back MCP Standard For Web-Surfing AI Agents; Google’s Sundar Pichai Asks A Shakespearean Question
Disclaimer: This content was partially produced with the help of Benzinga Neuro and was reviewed and published by Benzinga editors.
