OpenAI's GPT-5.4 now generally available in GitHub Copilot
OpenAI's GPT-5.4, an agentic coding model, is now generally available in GitHub Copilot. The model was tested on real-world software development scenarios and demonstrated improved coding capabilities.
GPT-5.4 — Quick Specs
OpenAI's GPT-5.4 Now Rolls Out to GitHub Copilot
OpenAI's GPT-5.4 is now generally available in GitHub Copilot, marking the expansion of the company's latest agentic coding model beyond initial testing phases.
What's New
GPT-5.4 joins GitHub Copilot's model lineup as an agentic model designed specifically for software development workflows. According to GitHub's announcement, the model was validated through early testing on real-world and agentic software development tasks, where it achieved improved performance rates compared to previous versions.
The exact context window size, pricing structure, and detailed benchmark scores for GPT-5.4 in this deployment have not been disclosed. GitHub's announcement referenced "new rates" achieved during testing but did not specify which metrics or benchmarks were used in evaluation.
Integration with GitHub Copilot
As part of GitHub Copilot, GPT-5.4 becomes available to both individual developers and enterprise users. The rollout appears to be gradual, with the model becoming generally accessible rather than limited to beta testers.
The model is positioned as an agentic system, meaning it can take multiple steps to solve coding problems autonomously, rather than simply suggesting completions or fixes. This represents a shift toward more complex, multi-step code generation and debugging tasks.
What This Means
GPT-5.4's availability in GitHub Copilot signals OpenAI's continued focus on enterprise developer tools and its competitive positioning against other AI coding assistants. The emphasis on agentic capabilities suggests GitHub and OpenAI are betting on multi-step reasoning and autonomous problem-solving as differentiators in the code generation market.
Developers using GitHub Copilot can now access this model, though tier availability and whether it requires separate subscriptions or comes standard with existing plans remains unclear from the announcement. Full technical specifications and comparative benchmarks are expected in forthcoming documentation.
Related Articles
Xiaomi Releases MiMo-V2.6-Flash-RL, a 309B-Parameter MoE Model with 1M-Token Context and Native Omnimodal Support
Xiaomi's MiMo team released MiMo-V2.6-Flash-RL, an efficiency-tier checkpoint in the MiMo-V2.6 series featuring a 309B-parameter (15B active) Mixture-of-Experts architecture, 1M-token context, and native support for text, image, video, and audio. The model uses a single mixed reinforcement learning run across coding, agentic, visual, and cybersecurity tasks rather than domain-specific training.
Robot Safety Benchmark Finds GPT-6 Astra and Claude Fable 5.1 Rarely Refuse Dangerous Commands
A new benchmark called RoboHarm tested whether AI models controlling robotic arms would refuse dangerous commands. GPT-6 Astra completed 60 of 100 dangerous tasks and Claude Fable 5.1 completed 34, with neither model showing a reliable safety layer.
Xiaomi Releases MiMo-V2.6-Pro-RL, a 1.02T-Parameter Omnimodal Model with 1M-Token Context
Xiaomi's MiMo team has released MiMo-V2.6-Pro-RL, a 1.02-trillion-parameter sparse mixture-of-experts model with 42B active parameters, 1M-token context, and native text/image/video/audio processing. The model was trained via a single mixed reinforcement learning run spanning coding, agentic, visual, and cybersecurity tasks, with benchmark scores that Xiaomi claims approach or match Claude Opus 5 and GPT-5.6 on several agentic and coding tests.
TypeSafe AI Launches Jev, a 'Decision Model' That Outputs Only Numbers, Priced at $0.042/M Input Tokens
TypeSafe AI has released Jev, the first model in a new category it calls 'System One models'—text goes in, floating-point decisions come out. At $0.042 per million input tokens with free output, it undercuts even GPT-5 Nano on price.
Comments
Loading...