You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/ai-for-process/release-notes/recent-updates.md
+79Lines changed: 79 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -2,6 +2,85 @@
2
2
3
3
This document provides information on the feature updates and enhancements introduced in the recent AI for Process releases.
4
4
5
+
## v1.0.9 February 24, 2026
6
+
7
+
<u> Minor Release </u>
8
+
9
+
This release introduces enhancements to AI-powered browser automation, workflow features, and general platform improvements. Key updates include the new Delay node, workflow caching, and browser automation enhancements such as configurable auto-healing, model connection updates, expanded model support, hardware upgrades, and improved audit logging.
10
+
11
+
<fontsize="4">Workflows</font>
12
+
13
+
**Delay Node for Workflows**
14
+
15
+
A new Delay Node allows users to pause workflow execution at specific steps. Users can configure the node name and define a timeout value, either static or variable-based, between 30 and 86,400 seconds. Workflow changes will be visible in debugging and monitoring tools.
16
+
17
+
**Workflow Cache Management**
18
+
19
+
Workflows now support a built-in cache for storing and retrieving values across steps or runs. Users can read and write cache data in Script nodes, use values in template-supported fields, choose account-level or workflow-level scope, and set custom TTLs. Draft and published runs maintain separate cache data.
20
+
21
+
**Browser Automation Enhancements:**
22
+
23
+
***Configurable Auto-Healing**
24
+
25
+
Users can now control LLM-based auto-healing through a dedicated toggle in the Browser node configuration. Auto-healing is no longer enabled by default for new workflows, while existing workflows keep it enabled for backward compatibility.
26
+
27
+
***Model Connection Configuration**
28
+
29
+
Users can now update the connection used by a model within a browser node after the automation is published. While the model itself cannot be changed, the associated connection can be switched to address issues such as expired API keys or rate limits.
30
+
31
+
***Import and Export Enhancements**
32
+
33
+
Browser automation nodes are now included in workflow exports. Imports do not apply auto-mapping, so components must match between environments. Before exporting, users should verify that all required components — such as models, linked workflows, custom scripts and functions, integrations, triggers, agentic apps, OCR models, and automations — exist in the target environment.
34
+
35
+
36
+
<fontsize="4">Models</font>
37
+
38
+
**Expanded Model Support for Browser Automation**
39
+
40
+
Browser automation now supports a wider set of LLMs across OpenAI, Anthropic, and Google. These additions give users more flexibility and access to the latest model capabilities. Supported models include:
41
+
42
+
OpenAI:
43
+
44
+
* gpt-4o
45
+
* gpt-4o-mini
46
+
* gpt-4.1
47
+
* gpt-5
48
+
49
+
Anthropic:
50
+
51
+
* claude-sonnet-4-20250514
52
+
* claude-sonnet-4-5-20250929
53
+
* claude-haiku-4-5-20251001
54
+
55
+
Google (Gemini):
56
+
57
+
* gemini-2.0-flash
58
+
* gemini-2.5-pro
59
+
* gemini-2.5-flash
60
+
* gemini-2.5-flash-lite
61
+
62
+
63
+
<fontsize="4">Settings</font>
64
+
65
+
**Browser Automation Enhancements:**
66
+
67
+
***Browser Automation Playground Locking**
68
+
69
+
The Browser Automation playground now prevents conflicting edits by allowing only one active editor at a time. When a user is editing an automation, all other users enter view-only mode, where editing, running instructions, updating models, saving drafts, and publishing are disabled. The playground unlocks when the active user leaves or after one hour of inactivity.
70
+
71
+
***Enhanced Hardware Options**
72
+
73
+
Users can now select larger hardware configurations to improve performance and increase parallel streaming capacity. In addition to the existing 4 vGPU / 16 GB option, an 8 vGPU / 32 GB configuration is now available. The platform also displays updated streaming limits based on the selected hardware, with credit consumption adjusted automatically.
74
+
75
+
***Expanded Overview Details for Published Automations**
76
+
77
+
The Overview page for published automations now displays detailed configuration information. Users can view key automation parameters, including system prompt, max steps, recording status, auto-healing state, Go to URL value, and file-upload settings, for improved transparency.
78
+
79
+
***Audit Log Enhancements**
80
+
81
+
Audit logs now capture additional account-level and workflow-level events related to browser automation, improving visibility into user actions and automation behavior across the platform.
Copy file name to clipboardExpand all lines: docs/ai-for-process/settings/billing/billing-and-usage.md
+33-11Lines changed: 33 additions & 11 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -16,11 +16,17 @@ The Usage page displays the following tabs:
16
16
***Workflow runs**: Shows current workflow run usage against your total allocation.
17
17
***Credits**: Displays the number of credits consumed out of the total available credits, along with the corresponding percentage. A dynamic pie chart visually represents this data, showing the distribution of total credits, including usage by models, guardrails, OCR models, and custom script deployments. Each metric is color-coded and identified through a legend for easy reference. Hover over the chart to view the actual values for each usage type.
18
18
***Usage trend**: Visual representation of workflow runs, model credits, custom scripts, and OCR credits consumption over time.
19
+
20
+
Click the "*three-dots*" icon to view and select from the following options:
21
+
19
22
***Workflows**: Shows all workflow activities and their operational status.
20
23
***Models**: Shows the computational costs linked to specific models.
21
24
***Guardrails**: Shows overall guardrail statistics and credit consumption costs.
22
25
***Custom Scripts**: Gives a comprehensive view of all the deployed custom scripts’ statistics, including the credits usage trends.
23
26
***OCR**: Displays OCR model deployments and their credit consumption, helping you track document-processing usage.
27
+
***Browser Automation**: Shows the credits consumption and usage trend for browser automation experiment pod deployments.
@@ -43,7 +49,7 @@ The **Overview** tab summarizes expenses and usage patterns across your workflow
43
49
44
50
## Workflows Usage
45
51
46
-
The Workflows tab displays a comprehensive list of workflows associated with the account. It includes only those workflows that have been deployed at least once; it does not include ‘In development’ workflows.
52
+
The **Workflows** tab displays a comprehensive list of workflows associated with the account. It includes only those workflows that have been deployed at least once; it does not include ‘In development’ workflows.
47
53
48
54
The following usage information is displayed on the tab:
49
55
@@ -55,7 +61,7 @@ The following usage information is displayed on the tab:
55
61
***Last active on**: The date when the workflow was last active.
56
62
***Status**: The workflow's status - Deployed, Undeployed, or Deleted.
@@ -93,15 +99,15 @@ The following information is displayed:
93
99
***Last updated on**: The date when the deployment was last done.
94
100
***Status**: The current deployment status: *Deployed*, *Undeployed*, or *Deleted*.
95
101
96
-
<imgsrc="../images/models-usage-drill-down.png"alt="drill down view"title="drill down view"style="border: 1pxsolidgray; zoom:80%;">
102
+
<imgsrc="../images/models-usage-drill-down.png"alt="drill down view"title="drill down view"style="border: 1pxsolidgray; zoom:80%;">
97
103
98
104
## Guardrails Usage
99
105
100
106
The **Guardrails** tab displays the list of guardrails used and the charges related to their usage. Each guardrail includes details about the hardware it was deployed on and the associated costs. The guardrails are charged based on the hardware used.
101
107
102
108
The guardrail usage is deducted from the Model credits shown in the Overview tab, indicating that the available credits for the model will decrease based on the cost of using the guardrails.
The following usage information is displayed on the Guardrails tab:
107
113
@@ -127,7 +133,7 @@ Clicking each row on the Guardrail tab opens a panel on the right that displays
127
133
* Further deployments are disabled.
128
134
* All ongoing deployments are disrupted once the negative credit limit is reached, and the message “*You've used all your available credits. Please add more credits to your account to continue.*” appears.
@@ -179,7 +185,7 @@ Clicking each row on the **Custom scripts** tab opens a panel on the right that
179
185
* Credits consumed by the components. Refer to the table [here](../manage-custom-scripts/custom-scripts.md#step-3-resource-allocation){:target="_blank"} for pricing.
180
186
***Total** which represents the aggregate of all the credit components (when multiple components are involved).
### Viewing Detailed Browser Automation Information
211
227
228
+
Clicking a row in the **Browser Automation** tab opens a right-side panel that displays detailed information about the selected OCR deployment, including:
212
229
230
+
***Hosting infrastructure:** The compute resources or allocated hardware used by the deployment.
231
+
***Hosting time:** Total time the deployment has been active, shown in hours and minutes.
232
+
***Credits consumed:** The number of browser automation credits used by the deployment.
233
+
***Total:** The aggregated credit usage across all contributing components for that deployment.
0 commit comments