> For the complete documentation index, see [llms.txt](https://documentation.virtuedeep.tech/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://documentation.virtuedeep.tech/whats-new/release-notes.md).

# Release notes

Learn about the new features and enhancements available in the current and previous releases of Affective Computing

## (Current version) 1.7.0

> Released on 16-Sep-2026

This release reshapes several core concepts. Three of the changes break existing integrations. Before upgrading, read [Migrating to the new model](https://documentation.virtuedeep.tech/whats-new/migrating-to-the-new-model.md), which covers what moved and what to change.

Nothing has been switched off. Superseded endpoints are hidden from the API reference but remain callable.

### One attribute, replacing three attribute types

Feeding content attributes, interaction attributes and triggered action attributes have been replaced by a single **attribute**. The type now belongs to the actions inside an attribute rather than to the attribute itself, so one attribute can hold media, interaction and triggered actions together.

A logic output names an action type and an attribute, and only the actions of that type inside it are eligible for delivery. The same attribute can therefore serve several outputs, each drawing the kind it asked for.

Five endpoint collections have been replaced by one. See [Working with attributes](https://documentation.virtuedeep.tech/guides/using-the-public-apis/working-with-actions-and-attributes/working-with-attributes.md).

### Feeds are tenant-level

Feeds belonged to a single project. They now belong to your account, and `projectId` has been removed from feed parameters and responses.

Project has not disappeared from feeding: you can still filter feeds and attributes by project, project domain or project function. The change is one of ownership, not visibility.

### Bring your own LLM

The platform no longer ships with a language model. Connect one to your account, either a custom HTTP endpoint or an AWS Bedrock model, before feeding anything.

Projects that hold a conversation with a real person also need a model assigned to them. Projects whose purpose is `Simulation` or `Simulation_bulk` run on virtual humans and are rejected if you try to assign one.

See [Configuring LLMs](https://documentation.virtuedeep.tech/guides/using-the-public-apis/configuring-llms.md).

### Solutions

A **solution** pairs one project domain with one project function and the model that runs them, as a catalogue entry the platform provides. Project creation now has two paths: start from a catalogue solution, which carries the domain, function and model, or supply the domain and function directly as before.

See [Working with solutions](https://documentation.virtuedeep.tech/guides/using-the-public-apis/working-with-solutions.md).

### Interaction channel outputs carry an identity

Interaction channels describe both directions. Inputs capture what comes into the platform; outputs deliver what it produces. Every output now carries an identity, so a consumer receiving several responses can tell which output produced each one without inferring it from the payload.

### Logics apply to many projects

A logic belonged to exactly one project. A logic can now apply to one or many, so a rule you want enforced across a portfolio is created once and attached to each project rather than duplicated.

A logic is scoped either by the projects attached to it or by a solution, never both.

### Blueprinting vocabulary

What was called a logic *condition* is now a **variable**, of one of three types: user motion, environmental and user geotargeting. What was called a logic *action* is now an **output**.

{% hint style="warning" %}
This is a naming change in the product and the documentation only. API request and response field names still say `condition` and `action`, and have not moved. They will be brought in line during a later API revision, which will be announced separately.
{% endhint %}

### Documentation

The API reference no longer embeds a live specification. Every page carries its own request and response tables, so the documentation stays available regardless of platform status. Pages state their full endpoint path individually rather than assuming a shared prefix.

## Previous releases

<table><thead><tr><th width="143">Version</th><th width="149">Release notes</th><th>Summary</th></tr></thead><tbody><tr><td>1.3.2</td><td><a data-mention href="https://documentation.virtuedeep.tech/whats-new/release-notes/1.3.2.3.md">1.3.2.3</a></td><td><ul><li>New <code>perceptions</code> response parameter on the Psychological Evolution API</li><li>Ability to share users and VHs across projects of the same tenant</li><li>New validations when summoning VHs</li></ul></td></tr><tr><td></td><td><a data-mention href="https://documentation.virtuedeep.tech/whats-new/release-notes/1.3.2.2.md">1.3.2.2</a></td><td><ul><li>Enhancements to the <code>Interaction/PsychologicalEvolution</code> API to provide clear indicators of Virtual Humans' psychological evolution</li><li>Tier-based data limits and maintenance protocols</li><li>Ability to create input interaction channels based on environmental objects</li><li>Additional validation and improved error handling</li><li>Support for retrying connection requests from gRPC clients</li></ul></td></tr><tr><td></td><td><a data-mention href="https://documentation.virtuedeep.tech/whats-new/release-notes/1.3.2.1.md">1.3.2.1</a></td><td><ul><li>Uniform API response structure</li><li>Improved error messaging</li><li>Support for a minimum of 2 participants in group interactions</li></ul></td></tr><tr><td>1.3.1</td><td><a data-mention href="https://documentation.virtuedeep.tech/whats-new/release-notes/1.3.1.1.md">1.3.1.1</a></td><td><ul><li><p>Support for group interactions:</p><ul><li>New <code>Interaction/group/InitializeInteraction</code> API for initializing group interactions</li><li>New <code>Interaction/group/DissolveInteraction</code> API for ending group interactions</li><li>New <code>group_ID</code> parameter in the <code>Interaction/Interact</code> API to implement group interactions</li><li>New <code>next_Participant_InteractionSession</code> parameter in the <code>Interaction/Interact</code> API to specify the next participant in directed group interactions</li></ul></li><li>New <code>Interaction/PsychologicalEvolution</code> API to gain insights about the psychological evolution of Virtual Humans during simulation.</li><li>New <code>interaction_identifier</code> response attribute in the <code>Interaction/Interact</code> API to identify successful interaction requests</li><li>Ability to manually verify actions using the new <code>FeedingData/VerifyActions</code> API</li><li>Validation of values entered in parameters when creating projects, logics, and actions and initializing interaction sessions.</li></ul></td></tr><tr><td></td><td><a data-mention href="https://documentation.virtuedeep.tech/whats-new/release-notes/1.3.1.md">1.3.1</a></td><td><ul><li>Enhancements to IAS to improve conversations based on action feeding, data augmentation, and third-party LLM integrations</li><li><p>Enhancements to actions</p><ul><li>All responses now pass grammar corrections before delivery</li><li>All actions are kept secret (delivered by EDAA™ only for logics defined within the project) by default</li></ul></li><li>New <code>Interaction/StartDataAugmentation_bulk</code> API to perform data augmentation for bulk simulation</li><li>New <code>project_Purpose</code> parameter in the <code>Project/Create</code> API to specify a new project's purpose</li><li>Ability to specify <code>simulation_bulk</code> as the project's purpose</li><li>Ability to create and manage diagnostics logics using the APIs in the new <code>logicDiagnostics</code> collection</li><li>Ability to enable or disable product calibration mode using the new <code>Project/ActivateProductCalibration</code> and <code>Project/DeactivateProductCalibration</code> APIs respectively.</li></ul></td></tr><tr><td>1.3.0</td><td><a data-mention href="https://documentation.virtuedeep.tech/whats-new/release-notes/1.3.0.md">1.3.0</a></td><td><ul><li>Removed the requirement to input physiological data in simulation projects</li><li>Ability to send text responses to EDAA™ (instead of only audio)</li><li>New (optional) <code>feedingDataId</code> parameter in the <code>Interaction/Interact</code> API to inform EDAA™ about actions that occur during data collection</li><li>New <code>Interaction/BroadcastAction</code> API to inform VHs about an action performed by another VH that impacts them.</li><li>EDAA™ only responds with logic actions in data collection projects.</li><li><p>Performance improvements:</p><ul><li>Fixed issues in summoning entities</li><li>Fixed issues with interactions by the environment in simulation projects</li><li>Improved the platform's overall response time</li></ul></li></ul></td></tr></tbody></table>
