Changelog

This section lists all of the features that have been removed, deprecated, added, or extended in different versions of the GenAI plugin.

Cypher 25 was introduced in Neo4j 2025.06 and can only be used on Neo4j 2025.06+ databases. For more information, see Cypher Manual → Select Cypher version.

Neo4j 2025.12

Updated in Cypher 25

Feature Details

Feature	Details
`RETURN ai.text.completion(prompt, provider, configuration)`	The configuration for `ai.text.completion` now accepts `chatHistory`, see the specific provider details for more information.

RETURN ai.text.completion(prompt, provider, configuration)

The configuration for ai.text.completion now accepts chatHistory, see the specific provider details for more information.

New in Cypher 25

Feature Details

Feature	Details
`RETURN ai.text.chat(prompt, chatId, provider, configuration)`	New function `ai.text.chat()` for conversational interactions that can continue prior responses. See Generate text → Chat with context.
`RETURN ai.text.chat.providers()`	New procedure `ai.text.chat.providers()` for information on the available providers for the `ai.text.chat()` function. See Generate text → Chat with context.
Support for Google VertexAI’s `apiKey`	The VertexAI provider now supports authentication via either `apiKey` or `token`. Applies to the following callables and their provider settings: `ai.text.completion` — see Generate text → Google VertexAI. `ai.text.embed` and `ai.text.embedBatch` — see Embeddings → Google VertexAI.
Custom base URL for OpenAI via the `genai.openai.baseurl` setting in the `genai.conf` file.	Applies to all `ai.text.*` calls that use OpenAI, including `ai.text.embed`, `ai.text.embedBatch` and `ai.text.completion`. See Configuration Options → genai.openai.baseurl.

RETURN ai.text.chat(prompt, chatId, provider, configuration)

New function ai.text.chat() for conversational interactions that can continue prior responses. See Generate text → Chat with context.

RETURN ai.text.chat.providers()

New procedure ai.text.chat.providers() for information on the available providers for the ai.text.chat() function. See Generate text → Chat with context.

Support for Google VertexAI’s apiKey

The VertexAI provider now supports authentication via either apiKey or token. Applies to the following callables and their provider settings:

ai.text.completion — see Generate text → Google VertexAI.
ai.text.embed and ai.text.embedBatch — see Embeddings → Google VertexAI.

Custom base URL for OpenAI via the genai.openai.baseurl setting in the genai.conf file.

Applies to all ai.text.* calls that use OpenAI, including ai.text.embed, ai.text.embedBatch and ai.text.completion. See Configuration Options → genai.openai.baseurl.

Neo4j 2025.11

New in Cypher 25

Feature Details

Feature	Details
`RETURN ai.text.completion(prompt, provider, configuration) CALL ai.text.completion.providers() YIELD name, requiredConfigType, optionalConfigType, defaultConfig`	New function `ai.text.completion()` for generating text with external AI providers and new procedure `ai.text.completion.providers()` to display the available providers and their config setup. See Generate text & Chat.
`RETURN ai.text.embed(resource, provider, configuration) CALL ai.text.embedBatch(resources, provider, configuration) YIELD index, vector CALL ai.text.embed.providers() YIELD name, requiredConfigType, optionalConfigType, defaultConfig`	New function `ai.text.embed` and procedure `ai.text.embedBatch` for generating vector embeddings. There is also a new procedure `ai.text.embed.providers()` to display the available providers and their config setup. See Create and store embeddings.

RETURN ai.text.completion(prompt, provider, configuration)

CALL ai.text.completion.providers() YIELD name, requiredConfigType, optionalConfigType, defaultConfig

New function ai.text.completion() for generating text with external AI providers and new procedure ai.text.completion.providers() to display the available providers and their config setup. See Generate text & Chat.

RETURN ai.text.embed(resource, provider, configuration)

CALL ai.text.embedBatch(resources, provider, configuration) YIELD index, vector

CALL ai.text.embed.providers() YIELD name, requiredConfigType, optionalConfigType, defaultConfig

New function ai.text.embed and procedure ai.text.embedBatch for generating vector embeddings. There is also a new procedure ai.text.embed.providers() to display the available providers and their config setup. See Create and store embeddings.

Deprecated in Cypher 25

Feature Details

Feature	Details
`RETURN genai.vector.encode(resource, provider, configuration) CALL genai.vector.encodeBatch(resources, provider, configuration)`	The function `genai.vector.encode()` and the procedure `genai.vector.encodeBatch()` are deprecated and replaced by `ai.text.embed()` and `ai.text.embedBatch()`.

RETURN genai.vector.encode(resource, provider, configuration)

CALL genai.vector.encodeBatch(resources, provider, configuration)

The function genai.vector.encode() and the procedure genai.vector.encodeBatch() are deprecated and replaced by ai.text.embed() and ai.text.embedBatch().