OpenAI Releases New Family of AI Models for Coding
OpenAI announced today that it is releasing a new family of artificial intelligence models optimized to excel at coding, as it ramps up efforts to fend off increasingly stiff competition from companies like Google and Anthropic. The models are available to developers through OpenAI’s application programming interface (API).
New Models and Features
OpenAI is releasing three sizes of models: GPT 4.1, GPT 4.1 Mini, and GPT 4.1 Nano. Kevin Weil, chief product officer at OpenAI, said on a livestream that the new models are better than OpenAI’s most widely used model, GPT-4o, and better than its largest and most powerful model, GPT-4.5, in some ways.
Performance and Capabilities
GPT-4.1 scored 55 percent on SWE-Bench, a widely used benchmark for gauging the prowess of coding models. The score is several percentage points above that of other OpenAI models. The new models are “great at coding, they’re great at complex instruction following, they’re fantastic for building agents,” Weil said.
Improved Capabilities
The new models can analyze eight times more code at once, which improves their ability to make improvements and fix bugs. The new models are also better at following instructions given by users, reducing the need to repeat commands in different ways to get the desired result.
Industry Reactions
Varun Mohan, CEO of Windsurf, a popular tool for AI coding, said that the company had been testing GPT-4.1 and found that the new model was “60 percent” better than GPT-4o according to its own benchmarks. “We found that GPT-4.1 has substantially fewer cases of degenerate behavior,” Mohan said, noting that the new model spends less time reading and editing irrelevant files by mistake.
Conclusion
The release of GPT-4.1 marks a significant step forward in OpenAI’s efforts to improve its AI models for coding. With its improved capabilities and performance, the new models are likely to be highly sought after by developers and companies looking to leverage AI for coding tasks.
FAQs
Q: What are the new AI models for coding?
A: OpenAI is releasing three sizes of models: GPT 4.1, GPT 4.1 Mini, and GPT 4.1 Nano.
Q: How do the new models perform?
A: GPT-4.1 scored 55 percent on SWE-Bench, a widely used benchmark for gauging the prowess of coding models.
Q: What are the improvements in the new models?
A: The new models can analyze eight times more code at once, which improves their ability to make improvements and fix bugs. They are also better at following instructions given by users.
Q: How does the new model compare to previous models?
A: GPT-4.1 is 40 percent faster than GPT-4o, OpenAI’s most widely used model for developers. The cost of users inputting queries has been reduced by 80 percent in this latest version.
