IDC Launches Quanta to Close Gap Between Tech Intelligence, Enterprise Execution
IDC has launched IDC Quanta to deliver intelligence inside the tools people use for work. Quanta, which the company describes as “the technology intelligence fabric of the AI-enabled enterprise,” offers IDC’s proprietary research, analysis data collected from 60+ years of research and billions of quantative data points.
“The ability to ask a question and get an insight backed by specific, referenceable research is really the key to the tool,” said Eric Walk, VP of AI and Data Platform Services, Perficient, a beta customer of IDC Quanta. “It’s become another tool in my tool belt that my main workbench — the one I work from every day — can make calls out to and return better research and better information to power what I’m doing.”
IDC Quanta pairs IDC intelligence with a customer’s own data and history. It allows customers to upload content, data, and context to be synthesized and harmonized alongside IDC’s verified insights, supporting more personalized, relevant responses, and an integrated view of data. Every response is verified against IDC’s proprietary data through a multi-agent system to validate qualitative research before responding. IDC Quanta leverages 15B data points to validate responses. Once responses are formulated, it is easy to understand how responses were generated with an expandable reasoning panel and full traceability. Citations identify the sources of insights, providing users with confidence in each response.
CircleCI Introduces Merge Efficiency Ratio
CircleCI has released new data from more than 20 million software delivery workflows that shows how development teams are using AI-generated code successfully. Its Q2 Pulse report introduce Merge Efficiency Ratio (MER), which the company said “tracks how many validation cycles it takes to ship one change.”
The company said the MER metrics connect the speed of delivering software with the cost, which it said matters more as the volume of AI-generated code increases in volume and becomes a bottleneck to organizations that must test and verify that code, not to mention that it must be secure. In its report, Circle CI found that 20 elite organizations it profiled grew their main-branch throughput 72% in a year while cutting cost per shipped change by 31%.
The full report is available here with no registration required.
OpenAI announces ChatGPT Work
The new ChatGPT agent can gather context from apps, files and workflows to create documents, spreadsheets and web apps. It’s powered by GPT-5.6, the company’s newest frontier model family, which includes Sol, Terra and Luna.
The way it works is that a unified plugins directory connects ChatGPT to the tools used in everyday work, such as Slack, Gmail, calendars and more. WIth Codex built in, ChatGPT moves from simply answering questions to get work done, even when users are away from their desktops or phones, through the user of new Scheduled Tasks.
Key performance markers include:
-
On Agents’ Last Exam, GPT‑5.6 Sol sets a new high of 53.6, eclipsing Claude Fable 5 (adaptive) by 13.1 points. At medium reasoning, it beats Fable 5 by 11.4 points at roughly one-quarter the estimated cost. GPT‑5.6 Terra and Luna also outperforms Fable 5 at around one-sixteenth the cost.
-
On the Artificial Analysis Coding Agent Index, GPT‑5.6 Sol sets a new state of the art at 80.0—2.8 points above Claude Fable 5—while using less than half the output tokens, taking less than half the time, and costing about one-third less.
-
On ExploitBench, a cybersecurity eval that measures progress from reaching vulnerable code through arbitrary code execution, it scores 73.5% versus GPT-5.5’s 47.9% at a comparable output-token budget.
-
On coding-agent tasks, GPT-5.6 Sol with max reasoning outperforms Claude Fable 5 while using 54% fewer output tokens. That advantage extends across the family: at their highest-scoring configurations, GPT-5.6 Terra edges Claude Fable 5 and GPT-5.6 Luna outperforms Opus 4.8, each at roughly one-fifth the estimated API cost and one-third of the time
-
On long-running professional workflows, GPT-5.6 Sol outperforms Fable by 9.1 points at roughly one-quarter the estimated API cost. Terra and Luna also outperform Fable at around one-sixteenth the cost.
