Claude Enterprise Merges Chat and Cowork: What Admins Control
Claude Enterprise now merges chat and Cowork into one Claude. What changes for users, the new Efficient mode and spend limits, and the Nov 16 default switch-on.
PromptWises is reader-supported. Some links may earn us a commission at no extra cost to you. This never affects our ratings or recommendations.
On October 23, Anthropic made the merged chat-and-Cowork experience generally available on Claude Enterprise, a month after it began rolling out to Pro and Max on September 16. Enterprise owners can switch it on now, for everyone or one role at a time. According to Anthropic’s admin guide, it turns on by default on November 16, 2026 unless an owner chooses to keep it off. For users it means one entry point for every task. For admins, the more important part is cost: agentic work uses more of your usage than chat, so the controls that come with it matter more than the merge.
What changes for Claude Enterprise users
There is no longer a choice between Chat and Cowork. Every conversation starts as chat. When a task needs code execution or file handling, Claude continues it in an Anthropic-managed cloud session and comes back with a finished document, analysis or deck. The work keeps running after you close your laptop, and you can pick it up again on web, desktop or mobile. Chats, projects, connectors, skills and plugins carry over.
The desktop app is still needed for anything local: local files, local connectors, browser use and computer use. Decks and documents inside chat came with the same September update; see our coverage of Claude Design moving into Claude, and our Claude review for how the plans compare.
How to turn it on (and the November 16 default)
An Owner or Primary Owner turns it on in Organization settings > Capabilities > “Chat and Cowork unified”. To limit it to some people, use Organization settings > Roles > (role) > Capabilities; the organisation-level setting must also be on. “Cloud code execution and file creation” has to be enabled first.
Until November 16, owners see a notice with two buttons: Enable now or Keep disabled. If someone has already changed the setting, that choice stays. If Chat or cloud code execution is off in your organisation, nothing turns on.
Anthropic’s suggested rollout is cautious: involve IT, security and procurement, pilot with one role or a sandbox organisation, review egress, connector, sharing and remote-control settings, watch it for two to four weeks, then expand in waves. SCIM can sync role membership for each wave.
Cost controls: Efficient mode, model limits and spend caps
Pricing stays usage-based. At the time of writing, Claude’s pricing page lists Enterprise at US$20 per seat per month billed annually, with a 20-seat minimum, plus usage that “scales with model and task”. There is no fixed per-user total, and cloud sessions use tokens faster than chat. Anthropic gives admins four levers:
| Control | What it does |
|---|---|
| Efficient mode (new) | Limits members to Sonnet and Haiku models; new sessions start on Claude Sonnet 5.5 at low effort. Set org-wide on the Models page or per role. |
| Model entitlements | One allowlist by role covers both chat and agentic work. |
| Spend limits | Caps at organisation, group and member level. |
| Pooled group budgets (beta) | One shared monthly amount per group; when it runs out, usage pauses for everyone in it. |
Efficient mode is a model setting, not a cheaper price tier. Anthropic also claims that Opus 5.5 costs 40% less to run than Opus 5 on typical workloads, and Sonnet 5.5 up to 30% less than Sonnet 5. These are vendor figures. For the models themselves, see our posts on Claude Sonnet 5.5 and Claude Haiku 5.5.
Security defaults and who has to wait
Each cloud session runs in its own isolated virtual machine. Network egress is off by default on Enterprise; admins can allow package managers only, a custom domain list or all domains. Connector credentials stay outside the sandbox, existing SSO and Compliance API controls apply, and data is not used for training by default.
Some organisations cannot turn it on yet: those using zero data retention, the HIPAA-ready configuration or customer-managed encryption keys (CMEK). Anthropic expects HIPAA support after general availability and plans CMEK for a later release; no dates are given. Turning on CMEK later makes the experience unavailable.
What changes in usage reports
Usage now appears under a new “Chat and Cowork unified” line, while past usage keeps its old label. Counting and billing do not change, but finance teams totalling chat and Cowork spend need to add the new line. In the Cost and Usage API the product value is chat_cowork_unified, and new activity can take up to two days to show. On retention, sandbox environments expire 30 days after creation, and cloud sessions started from a conversation are deleted along with it.
Claude Enterprise is not the only way companies buy Claude; Anthropic-managed Claude is also sold through Google Cloud, which we covered in OpenAI and Claude on Google Cloud. More updates are in our news section.
Frequently asked questions
What is the difference between Claude chat and Cowork on Enterprise now?
Once an owner turns on the "Chat and Cowork unified" setting, there is nothing to choose. Every conversation starts as chat, and when a task needs code execution or file handling, Claude continues it in an Anthropic-managed cloud session and returns the finished document, analysis or deck.
When will the unified Claude experience turn on for my Enterprise organisation?
Owners can enable it now, for everyone or by role. In existing organisations it is off by default and turns on automatically on November 16, 2026, unless an owner selects "Keep disabled". If the setting was already changed, that choice is kept.
What is Efficient mode in Claude Enterprise?
An admin setting that limits members to Sonnet and Haiku models in the unified experience. New sessions start on Claude Sonnet 5.5 at low effort. Owners can switch it on for the whole organisation on the Models page or for a custom role. It does not affect other Claude products.
Does the unified Claude experience cost more?
Pricing does not change and stays usage-based, but agentic work in the cloud uses more tokens than a quick chat. Anthropic points admins to Efficient mode, model limits by role, spend limits at organisation, group and member level, and pooled group budgets.
Can HIPAA or zero-data-retention organisations use it?
Not yet. Organisations using zero data retention, the HIPAA-ready configuration or customer-managed encryption keys cannot turn it on and stay on their current setup. Anthropic says HIPAA support is expected after general availability and CMEK support is planned for a later release.
