OpenAI Enables Voice Commands to Automate Daily Tasks Across Workplace Apps
San Francisco, Wednesday, 23 September 2026.
OpenAI upgraded ChatGPT Voice with GPT-6 technology, enabling hands-free task execution across core workplace tools like email and Slack to streamline daily business operations.
OpenAI Deploys GPT-6 Voice Integration Across Enterprise Platforms
OpenAI has officially upgraded ChatGPT Voice to integrate directly with enterprise workspace tools, marking a significant shift in office automation technology. Announced globally on 22 September 2026, the update powers the voice interface with the latest GPT-6 models, including Astra, Sol, and Luna [1][2]. This deployment allows enterprise users to execute hands-free voice commands across core platforms such as Slack, email, and calendars on both web and mobile interfaces [1][3]. The update aims to transform conversational AI into a central interface for workflow management, moving beyond simple queries to actionable task execution [2][6].
Model Architecture and Pricing Structure
The upgraded system utilizes three distinct GPT-6 variants tailored for different computational needs. GPT-6 Astra serves as the most capable model for complex workflows, while Sol and Luna offer medium and small intelligence configurations respectively [1][5]. API pricing for GPT-6 Sol is set at $2 per million input tokens and $10 per million output tokens [3]. In contrast, GPT-6 Luna is priced at $0.10 for input and $0.50 for output tokens, representing a significant cost differential where Sol’s input cost is 20 times higher than Luna’s [3]. These models are accessible through ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, and Edu plan subscribers [3][5].
Distinction Between Chat and Work Modes
A critical component of this rollout is the functional separation between standard Chat and the new Work modes. While Chat facilitates collaborative, step-by steering, Work is designed for delegation, allowing the AI to autonomously manage multi-step tasks and process files [4]. Despite identical user interfaces featuring a prompt bar, users are advised to utilize Work mode for larger tasks involving multiple files or moving parts [4][5]. This distinction clarifies previous confusion regarding model availability, as GPT-6 variants are primarily integrated into the enterprise-focused Work tools rather than the standard Chat interface [4][5].
Operational Limitations and Market Reception
Early feedback highlights potential workflow limitations despite the technical upgrades. Critics note that while voice execution sounds efficient, the integration may lack specific organizational context, potentially increasing correction time compared to manual tasks [1][5]. Some users observe that without clear prioritization logic, the tool solves the interface but not the real flow of enterprise work [1]. However, proponents argue that talking through tasks while the AI operates across browsers and documents turns voice into a viable workspace operator [1]. The rollout status as of 23 September 2026 indicates global availability in the latest app version [1][6].