OpenAI introduces BioBasicsBench for measuring AI agent performance in wet-lab science
New benchmark available for evaluating AI agents' ability to handle scientific research tasks, relevant to researchers and scientific organizations.
Evidence
+Measuring how AI agents handle everyday wet-lab science
OpenAI introduces BioBasicsBench, a 100-task benchmark measuring how well AI agents interpret raw experimental data and handle everyday wet-lab science.
OpenAI highlighting educational adoption and AI-native graduate preparation in Australian universities
Demonstrates growing adoption in education sector and positions OpenAI tools as integral to workforce development.
Evidence
+Australian universities are creating AI-native graduates
Australian universities are expanding access to ChatGPT and bringing AI into teaching and learning to prepare students for an AI-enabled workforce.
OpenAI publishes court brief excerpts defending fair use for AI training
OpenAI is publicly positioning itself as supportive of fair use principles in AI training, sharing evidence from creators, academics, and legal experts to support this stance.
Evidence
+Fair use benefits everyone
OpenAI shares excerpts from court briefs by creators, academics, accessibility advocates, and legal experts, among others, on how fair use supports AI training and innovation.
OpenAI publishes white paper on frontier AI governance for cyber defense
OpenAI is advocating for specific approaches to governing frontier AI with focus on cybersecurity, including safeguards and modernized defenses for public institutions.
Evidence
+Democratic governance of frontier AI, part II: Empowering cyber defenders
OpenAI outlines a blueprint for empowering cyber defenders with frontier AI, stronger safeguards, modernized cybersecurity, and more resilient public institutions.
OpenAI-hosted sandbox compute now billed only during active turns, not idle time
Developers reduce costs by eliminating charges for idle time between turns in Agents API sessions; billing is now more efficient and only covers active computation periods.
Evidence
+OpenAI-hosted sandbox compute is now billed during active turns. You no longer pay for idle time between turns.
Added Ultrafast mode for GPT-6.1 Sol in the Responses API
API users can now use the gpt-6.1-sol model with service_tier: "ultrafast" to reduce latency between output tokens, available globally with US and EU data residency options
Evidence
+Added Ultrafast mode for GPT-6.1 Sol in the Responses API. Use gpt-6.1-sol with service_tier: "ultrafast" to reduce the time between generated output tokens. It is available to all API users, subject to rate limits, with global processing and US and EU data residency.
OpenAI expands Atlassian partnership to integrate frontier models with enterprise knowledge management
Reinforces OpenAI's strategy of embedding AI capabilities into enterprise software platforms, enabling teams to leverage AI for knowledge-driven work planning and delivery.
Evidence
+Atlassian and OpenAI expand partnership to turn enterprise knowledge into action
Atlassian and OpenAI are expanding their partnership to connect frontier models with enterprise knowledge and help teams plan, build, and deliver work.
Pricing announced for GPT-6 Sol and GPT-6 Luna models
Users can now see cost differences between the two models: GPT-6 Sol has higher pricing ($2 input/$10 output) vs GPT-6 Luna ($0.10 input/$0.50 output), with both offering cached input options.
Evidence
+Standard pricing per 1M tokens for prompts with up to 272K input tokens:
GPT-6 Sol: $2 input, $0.20 cached input, and $10 output.
GPT-6 Luna: $0.10 input, $0.01 cached input, and $0.50 output.
Clarified that expanded input_file support applies to the Responses API, not v1/chat/completions
Developers need to understand that the new input_file type support is available in the Responses API specifically. This affects how developers should implement file handling in their applications.
Evidence
−v1/chat/completions
Expanded input_file support to accept more document, presentation, spreadsheet, code, and text file types.
+MCP servers
Expanded input_file support in the Responses API to accept more document, presentation, spreadsheet, code, and text file types.
Added API key governance controls at organization and project levels
Administrators can now enforce stricter API key management policies, restricting key creation to service accounts or project keys, and existing keys remain unaffected. This improves security and compliance for organizations.
Evidence
+Added API key creation governance controls at the organization and project levels. Administrators can allow only service-account keys, allow only user-owned project keys, or disable all new API key creation. Organization restrictions take precedence over project settings, and existing API keys are unaffected.
Prompt Cache Diagnostics now generally available in Responses API for GPT-5.6+
Developers using the Responses API can now diagnose and optimize prompt caching behavior, helping identify cache misses and improve reuse efficiency.
Evidence
+Prompt Cache Diagnostics is now generally available in the Responses API for GPT-5.6 and later supported models.
Compare cache reuse against a previous response, identify reasons for cache misses, and follow troubleshooting guidance to improve cache reuse.
OpenAI expands journalism support initiatives across education and news sectors
OpenAI is broadening its commitment to journalism by providing tools, training, and partnerships for students, educators, journalists, and news organizations.
Evidence
+OpenAI expands initiatives to support journalism from classrooms to newsrooms
OpenAI is expanding support for journalism with tools, training, and partnerships for students, educators, journalists, and news organizations.
ChatGPT for Teachers expansion to 55 U.S. school districts
Educators and staff: Over 100,000 more educators and staff across 55 U.S. school systems now have access to secure AI tools, training, and support through ChatGPT for Teachers.
Evidence
+ChatGPT for Teachers is expanding to 55 U.S. school systems, bringing secure AI tools, training, and support to over 100,000 more educators and staff.
GPT-5.6 Sol pricing reduced with 20% lower input and 33% lower output costs
API customers using GPT-5.6 Sol will see significant cost savings, with input tokens at $4/M and output tokens at $20/M through at least November 21, 2026.
Evidence
+GPT-5.6 Sol now costs $4 per million input tokens and $20 per million output tokens, representing 20% lower input pricing and 33% lower output pricing. GPT-5.6 Sol's promotional pricing is available at least through November 21, 2026.
ChatGPT Enterprise highlighted for enterprise productivity and advisory services
HSP GRUPPE case study demonstrates ChatGPT Enterprise use cases in professional services (tax advisory), positioning it as a productivity tool for knowledge workers
Evidence
+How HSP GRUPPE builds AI capabilities for tax advisory
Discover how HSP GRUPPE uses ChatGPT Enterprise to boost productivity, improve work quality, and create more capacity for tax advisory and client service.
Fast mode now supports long-context requests (272K+ tokens) for GPT-5.6 models with 2.5× speed improvement
Developers using GPT-5.6 Sol, Terra, or Luna can now use Fast mode for long-context prompts exceeding 272K tokens, achieving significantly faster response times.
Evidence
+Fast mode now supports long-context requests for GPT-5.6 Sol, GPT-5.6 Terra, and GPT-5.6 Luna. As of today, long-context prompts exceeding 272K tokens can run in Fast mode, delivering speeds up to 2.5× faster than the Standard tier.
Priority processing renamed to Fast mode with updated performance positioning
API users using priority processing will see their requests automatically migrated to the new Fast mode offering; branding change from 'Priority processing' to 'Fast mode'
Evidence
−Priority processing
+Fast mode
We're also introducing Fast mode in the API, which replaces our Priority Processing offering.
Developers have access to new API endpoints v1/audio/transcriptions and v1/realtime for integrating transcription capabilities into their applications.
OpenAI releases field report on AI coding agents for scientific computing
Scientists and researchers can learn how AI agents are being applied to accelerate genomics research and scientific software development workflows.
Evidence
+Scientific computing in the age of agentic AI
A new field report shows how scientists use AI coding agents to modernize scientific computing, accelerating software development and discovery in genomics and beyond.