Why Claude Opus 4.6 Is Changing How Enterprises Use AI

Anthropic’s latest release, Claude Opus 4.6, reflects a shift that goes beyond model upgrades or benchmark wins. The focus is increasingly on whether AI systems can be trusted to handle sustained, real-world work inside enterprise environments.

Opus 4.6 is designed to plan more carefully, sustain agentic tasks for longer, operate reliably across large codebases, and improve its own code review and debugging. These capabilities point to a clear direction: AI is moving from short, assistive interactions towards longer-running, semi-autonomous workflows.

Claude Opus 4.6 builds on its predecessor with stronger reasoning, better judgement in ambiguous situations, and the ability to remain productive over extended sessions. This matters for enterprise teams where work unfolds across hours or days, not single prompts.

According to Anthropic, the model brings more focus to complex parts of a task without instruction, moves faster through simpler steps, and revisits its own reasoning before finalising outputs. While deeper reasoning can introduce latency or cost on simpler jobs, developers can tune this behaviour using effort controls.

This balance between depth and efficiency is central to making AI viable in production settings.

Long context addresses a real enterprise pain point

A notable addition is the 1-million-token context window, currently in beta, the first for an Opus-class model. This enables Claude to work across significantly larger documents, datasets, and repositories without losing track of earlier details.

“Context rot”, where model performance degrades as conversations grow longer, has been a persistent concern for enterprises. Anthropic says Opus 4.6 performs markedly better in long-context retrieval and reasoning, allowing it to surface buried information and maintain coherence across complex workflows.

One early example comes from Rakuten. As shared by Yusuke Kaji, General Manager, AI, Rakuten:

“Claude Opus 4.6 autonomously closed 13 issues and assigned 12 issues to the right team members in a single day, managing a ~50-person organization across 6 repositories. It handled both product and organizational decisions while synthesizing context across multiple domains, and knew when to escalate to a human.”

The example illustrates how enterprises are beginning to test AI autonomy within live operational environments—while still keeping humans in the loop.

Anthropic is also introducing agent teams in Claude Code as a research preview. Instead of a single AI working sequentially, multiple agents can now operate in parallel, coordinating tasks in a way that reflects real engineering team structures.

Mario Rodriguez, Chief Product Officer, GitHub, described early results:

“Early testing shows Claude Opus 4.6 delivering on the complex, multi-step coding work developers face every day—especially agentic workflows that demand planning and tool calling. This starts unlocking long horizon task at the frontier.”

For enterprises managing large codebases, this approach reduces the need for constant human supervision while preserving structured execution.

Autonomy without losing developer control

Greater autonomy raises questions around cost, reliability, and safety. To address this, Anthropic has added adaptive thinking and expanded effort controls, allowing developers to choose how deeply the model reasons depending on the task.

Context compaction, now in beta, automatically summarises older parts of a conversation to keep long-running workflows active without hitting context limits.

Austin Ray, Staff Software Engineer, Ramp, highlighted how this changes day-to-day use:

“Claude Opus 4.6 is the biggest leap I've seen in months. I'm more comfortable giving it a sequence of tasks across the stack and letting it run. It's smart enough to use subagents for the individual pieces.”

Handling large codebases with confidence

For teams working across complex software environments, navigating large repositories remains a critical challenge. Early users say Opus 4.6 handles this more reliably than earlier models.

Amritansh Raghav, Interim CTO, Asana, noted:

“Claude Opus 4.6 felt like a clear step up. Code, reasoning, and planning were excellent. Its ability to navigate a large codebase and identify the right changes feels state of the art.”

This aligns with Anthropic’s emphasis on sustained reasoning and long-horizon execution rather than short-term output optimisation.

Extending beyond engineering into knowledge work

While coding remains central, Anthropic is also pushing Claude deeper into everyday enterprise tools. Updates to Claude in Excel improve its ability to handle unstructured data, plan multi-step actions, and execute changes in a single pass. Claude in PowerPoint, now in research preview, focuses on generating and restructuring presentations while respecting existing layouts and brand guidelines.

Yashodha Bhavnani, Head of AI, Box, shared results from internal evaluations:

“Claude Opus 4.6 excels in high-reasoning tasks, like multi-source analysis, across legal, financial, and technical content. Box’s eval showed a 10% lift in performance, reaching 68% vs. a 58% baseline, and near-perfect scores in technical domains.”

Similarly, design and product teams are exploring its creative potential. Loredana Crisan, Chief Design Officer, Figma, said:

“Claude Opus 4.6 generates complex, interactive apps and prototypes in Figma Make with an impressive creative range. The model translates detailed designs and multi-layered tasks into code on the first try, making it a powerful starting point for teams to explore and build ideas.”

Sustaining focus on harder problems

Across multiple customer accounts, one theme recurs: endurance. Enterprises are less interested in speed alone and more focused on whether AI systems can stay engaged when problems become messy or prolonged.

Michael Truell, co-founder & CEO, Cursor, summed up this shift:

“Claude Opus 4.6 stands out on harder problems. Stronger tenacity, better code review, and it stays on long-horizon tasks where others drop off. The team is really excited.”

Safety positioned as a prerequisite

Anthropic says these gains do not come at the expense of safety. Opus 4.6 reportedly maintains a safety profile as strong as, or better than, previous frontier models, with low rates of misaligned behaviour such as deception or over-refusal.

The company has expanded evaluations and added new cybersecurity probes, particularly as the model shows stronger capabilities in diagnosing software failures. The aim, Anthropic says, is to ensure that increased autonomy is paired with robust safeguards.

Claude Opus 4.6 is less about headline benchmarks and more about a shift in enterprise behaviour. Organisations are beginning to trust AI systems with longer-running, higher-stakes work, while expecting them to know when to escalate to humans.

For enterprises experimenting with AI at scale, Opus 4.6 represents a step towards AI that functions as operational infrastructure rather than an experimental tool.