genkit-ai · xavidop · Jul 6, 2026 · Jul 6, 2026 · Jul 6, 2026 · Jul 6, 2026
diff --git a/.gitignore b/.gitignore
@@ -31,4 +31,6 @@ node_modules
 hs_err_pid*
 samples/google-genai/generated_media/
 .astro
-.snapshots
+.snapshots
+
+samples/milvus/volumes
diff --git a/README.md b/README.md
@@ -438,7 +438,7 @@ Genkit genkit = Genkit.builder()
         .addRetriever(FirestoreRetrieverConfig.builder()
             .name("my-docs")
             .collection("documents")
-            .embedderName("googleai/text-embedding-004")
+            .embedderName("googleai/gemini-embedding-001")
             .vectorField("embedding")
             .contentField("content")
             .distanceMeasure(FirestoreRetrieverConfig.DistanceMeasure.COSINE)
@@ -476,7 +476,7 @@ public class MyFunction implements HttpFunction {
 
         genkit.defineFlow("generatePoem", String.class, String.class, (ctx, topic) -> {
             return genkit.generate(GenerateOptions.builder()
-                .model("googleai/gemini-2.0-flash")
+                .model("googleai/gemini-2.5-flash")
                 .prompt("Write a poem about: " + topic)
                 .build()).getText();
         });

diff --git a/docs/astro.config.mjs b/docs/astro.config.mjs
@@ -142,6 +142,10 @@ export default defineConfig({
             { label: "Weaviate", slug: "plugins/weaviate" },
             { label: "PostgreSQL (pgvector)", slug: "plugins/postgresql" },
             { label: "Pinecone", slug: "plugins/pinecone" },
+            { label: "Chroma", slug: "plugins/chroma" },
+            { label: "Qdrant", slug: "plugins/qdrant" },
+            { label: "Milvus", slug: "plugins/milvus" },
+            { label: "MongoDB", slug: "plugins/mongodb" },
           ],
         },
         {

diff --git a/docs/src/content/docs/agents/session-stores.md b/docs/src/content/docs/agents/session-stores.md
@@ -1,6 +1,6 @@
 ---
 title: Session Stores
-description: Persist agent session state with in-memory, file-based, Firestore, DynamoDB, or Cosmos DB session stores.
+description: Persist agent session state with in-memory, file-based, Firestore, DynamoDB, Cosmos DB, PostgreSQL, or MongoDB session stores.
 ---
 
 A `SessionStore<S>` is where a server-managed agent keeps its sessions. Pass one to `.store(...)` when you define an agent to persist snapshots; omit it to run the agent client-managed (stateless). Several implementations ship out of the box, and you can write your own.
@@ -12,8 +12,10 @@ A `SessionStore<S>` is where a server-managed agent keeps its sessions. Pass one
 | `FirestoreSessionStore` | Google Cloud Firestore | Production on Google Cloud (Cloud Run, Firebase Functions) |
 | `DynamoDbSessionStore` | Amazon DynamoDB | Production on AWS |
 | `CosmosSessionStore` | Azure Cosmos DB | Production on Azure |
+| `PostgresSessionStore` | PostgreSQL | Production on any SQL-first stack |
+| `MongoSessionStore` | MongoDB | Production on any document-first stack |
 
-If you plan to use `chat.abort()` or the `/abort` HTTP endpoint, pick a store that supports change notifications: `FileSessionStore`, `FirestoreSessionStore`, `DynamoDbSessionStore`, and `CosmosSessionStore` do; `InMemorySessionStore` does not, so `abort()` is a no-op there. See [Sessions](../sessions#aborting-a-turn).
+If you plan to use `chat.abort()` or the `/abort` HTTP endpoint, pick a store that supports change notifications: `FileSessionStore`, `FirestoreSessionStore`, `DynamoDbSessionStore`, `CosmosSessionStore`, `PostgresSessionStore`, and `MongoSessionStore` do; `InMemorySessionStore` does not, so `abort()` is a no-op there. See [Sessions](../sessions#aborting-a-turn).
 
 ## InMemorySessionStore
 
@@ -199,6 +201,100 @@ CosmosSessionStore<Map<String, Object>> store =
 
 The default shard size is 1 MiB, kept under Cosmos DB's 2 MB document-size limit. `CosmosSessionStore` supports `onSnapshotStateChange` (via polling), so `chat.abort()` works.
 
+## PostgresSessionStore
+
+Stores snapshots in PostgreSQL. It lives in the PostgreSQL plugin, so add the dependency:
+
+```xml
+<dependency>
+    <groupId>com.google.genkit</groupId>
+    <artifactId>genkit-plugin-postgresql</artifactId>
+    <version>${genkit.version}</version>
+</dependency>
+```
+
+```java
+import com.google.genkit.plugins.postgresql.session.PostgresSessionStore;
+import com.google.genkit.plugins.postgresql.session.PostgresSessionStoreOptions;
+import org.postgresql.ds.PGSimpleDataSource;
+
+PGSimpleDataSource dataSource = new PGSimpleDataSource();
+dataSource.setUrl("jdbc:postgresql://localhost:5432/genkit");
+dataSource.setUser("postgres");
+dataSource.setPassword("postgres");
+
+PostgresSessionStore<Map<String, Object>> store =
+    new PostgresSessionStore<>(dataSource); // uses the default "genkit_sessions" table
+
+Agent<Map<String, Object>> agent = genkit.beta().defineAgent(
+    AgentConfig.<Map<String, Object>>builder()
+        .name("myAgent")
+        .system("You are a helpful assistant.")
+        .store(store)
+        .build());
+```
+
+All records live in a single table keyed by `(pk, id)` with a JSONB `doc` payload and a `version` column for optimistic concurrency. Create the table ahead of time, or let the store create it on first use:
+
+```java
+PostgresSessionStore<Map<String, Object>> store =
+    new PostgresSessionStore<>(
+        dataSource,
+        PostgresSessionStoreOptions.builder()
+            .tableName("genkit_sessions")     // table name (default "genkit_sessions")
+            .createTableIfNotExists(true)     // auto-create the table on first use (default false)
+            .pollIntervalMs(500)              // how often change subscribers are polled (default 2000)
+            .build());
+```
+
+The default shard size is 1 MiB. `PostgresSessionStore` supports `onSnapshotStateChange` (via polling), so `chat.abort()` works.
+
+## MongoSessionStore
+
+Stores snapshots in MongoDB. It lives in the MongoDB plugin, so add the dependency:
+
+```xml
+<dependency>
+    <groupId>com.google.genkit</groupId>
+    <artifactId>genkit-plugin-mongodb</artifactId>
+    <version>${genkit.version}</version>
+</dependency>
+```
+
+```java
+import com.google.genkit.plugins.mongodb.session.MongoSessionStore;
+import com.google.genkit.plugins.mongodb.session.MongoSessionStoreOptions;
+import com.mongodb.client.MongoClient;
+import com.mongodb.client.MongoClients;
+
+MongoClient client = MongoClients.create("mongodb://localhost:27017");
+
+MongoSessionStore<Map<String, Object>> store =
+    new MongoSessionStore<>(client); // uses database "genkit", collection "genkit_sessions"
+
+Agent<Map<String, Object>> agent = genkit.beta().defineAgent(
+    AgentConfig.<Map<String, Object>>builder()
+        .name("myAgent")
+        .system("You are a helpful assistant.")
+        .store(store)
+        .build());
+```
+
+All records live in a single collection whose `_id` is `<prefix>::<recordId>`; each document carries a `version` field for optimistic concurrency. The database and collection are created automatically on first write. Tune it with `MongoSessionStoreOptions`:
+
+```java
+MongoSessionStore<Map<String, Object>> store =
+    new MongoSessionStore<>(
+        client,
+        MongoSessionStoreOptions.builder()
+            .databaseName("genkit")           // database name (default "genkit")
+            .collectionName("genkit_sessions") // collection name (default "genkit_sessions")
+            .pollIntervalMs(500)              // how often change subscribers are polled (default 2000)
+            .build());
+```
+
+The default shard size is 1 MiB, kept under MongoDB's 16 MB document-size limit. `MongoSessionStore` supports `onSnapshotStateChange` (via polling), so `chat.abort()` works.
+
 ## Writing your own store
 
 Implement `SessionStore<S>` to use any backend — a database, a cache, a cloud object store:

diff --git a/docs/src/content/docs/getting-started.mdx b/docs/src/content/docs/getting-started.mdx
@@ -186,7 +186,7 @@ Genkit genkit = Genkit.builder()
 
 ModelResponse response = genkit.generate(
     GenerateOptions.builder()
-        .model("googleai/gemini-2.0-flash")
+        .model("googleai/gemini-2.5-flash")
         .prompt("Tell me a fun fact!")
         .build());
 ```

diff --git a/docs/src/content/docs/index.mdx b/docs/src/content/docs/index.mdx
@@ -214,6 +214,36 @@ The fastest way to build AI features into your Java apps.
   </TabItem>
 </Tabs>
 
+## Build stateful agents
+
+Define server-managed agents with memory, tools, and multi-agent delegation. Persist conversations in the store of your choice.
+
+```java
+import com.google.genkit.agent.AgentConfig;
+import com.google.genkit.ai.agent.*;
+
+Agent<Map<String, Object>> assistant = genkit.beta().defineAgent(
+    AgentConfig.<Map<String, Object>>builder()
+        .name("assistant")
+        .system("You are a helpful assistant. Remember what the user tells you.")
+        .model("googleai/gemini-3-flash")
+        .store(sessionStore) // persist sessions in Firestore, DynamoDB, Cosmos DB, PostgreSQL, MongoDB, ...
+        .build());
+
+AgentChat<Map<String, Object>> chat = assistant.chat();
+chat.send("My name is Ada Lovelace.");
+String reply = chat.send("What is my name?").text(); // "Your name is Ada Lovelace."
+```
+
+<CardGrid>
+  <LinkCard title="Agents Overview" href="/genkit-java/agents/overview/" />
+  <LinkCard title="Define Agents" href="/genkit-java/agents/define-agents/" />
+  <LinkCard title="Sessions" href="/genkit-java/agents/sessions/" />
+  <LinkCard title="Session Stores" href="/genkit-java/agents/session-stores/" />
+  <LinkCard title="Multi-agent Delegation" href="/genkit-java/agents/multi-agent-delegation/" />
+  <LinkCard title="Serve over HTTP" href="/genkit-java/agents/serve-over-http/" />
+</CardGrid>
+
 ## Supported Providers
 
 <CardGrid>
@@ -238,6 +268,10 @@ The fastest way to build AI features into your Java apps.
   <LinkCard title="Weaviate" href="/genkit-java/plugins/weaviate/" />
   <LinkCard title="PostgreSQL (pgvector)" href="/genkit-java/plugins/postgresql/" />
   <LinkCard title="Pinecone" href="/genkit-java/plugins/pinecone/" />
+  <LinkCard title="Chroma" href="/genkit-java/plugins/chroma/" />
+  <LinkCard title="Qdrant" href="/genkit-java/plugins/qdrant/" />
+  <LinkCard title="Milvus" href="/genkit-java/plugins/milvus/" />
+  <LinkCard title="MongoDB (Atlas Vector Search)" href="/genkit-java/plugins/mongodb/" />
   <LinkCard title="Local Vector Store" href="/genkit-java/plugins/localvec/" />
 </CardGrid>
 

diff --git a/docs/src/content/docs/models.mdx b/docs/src/content/docs/models.mdx
@@ -59,7 +59,7 @@ genkit.generate(GenerateOptions.builder()
 
 // Use Gemini
 genkit.generate(GenerateOptions.builder()
-    .model("googleai/gemini-2.0-flash")
+    .model("googleai/gemini-2.5-flash")
     .prompt("Hello!").build());
 
 // Use Claude

diff --git a/docs/src/content/docs/plugins/anthropic.md b/docs/src/content/docs/plugins/anthropic.md
@@ -32,25 +32,26 @@ Genkit genkit = Genkit.builder()
 
 ModelResponse response = genkit.generate(
     GenerateOptions.builder()
-        .model("anthropic/claude-sonnet-4-5-20250929")
+        .model("anthropic/claude-sonnet-5")
         .prompt("Tell me about AI")
         .build());
 ```
 
 ## Available models
 
-### Claude 4.5 family
+### Claude 5 family
+- `anthropic/claude-fable-5`
+- `anthropic/claude-sonnet-5`
+
+### Claude 4 family
+- `anthropic/claude-opus-4-8`
+- `anthropic/claude-opus-4-7`
+- `anthropic/claude-opus-4-6`
+- `anthropic/claude-sonnet-4-6`
 - `anthropic/claude-opus-4-5-20251101`
 - `anthropic/claude-sonnet-4-5-20250929`
 - `anthropic/claude-haiku-4-5-20251001`
-
-### Claude 4 family
-- `anthropic/claude-4-*`
-
-### Claude 3 family
-- `anthropic/claude-3-opus-*`
-- `anthropic/claude-3-sonnet-*`
-- `anthropic/claude-3-haiku-*`
+- `anthropic/claude-opus-4-1-20250805`
 
 ## Features
 

diff --git a/docs/src/content/docs/plugins/aws-bedrock.md b/docs/src/content/docs/plugins/aws-bedrock.md
@@ -51,6 +51,19 @@ ModelResponse response = genkit.generate(
 - **OpenAI** — Via Bedrock marketplace
 - And many more...
 
+## Embeddings
+
+Bedrock embedding models (Amazon Titan and Cohere Embed) are available via the `InvokeModel` API:
+
+- `aws-bedrock/amazon.titan-embed-text-v2:0`
+- `aws-bedrock/cohere.embed-english-v3`
+- `aws-bedrock/cohere.embed-multilingual-v3`
+
+```java
+EmbedResponse response =
+    genkit.embed("aws-bedrock/amazon.titan-embed-text-v2:0", documents);
+```
+
 ## Features
 
 - Multi-provider model access

diff --git a/docs/src/content/docs/plugins/chroma.md b/docs/src/content/docs/plugins/chroma.md
@@ -0,0 +1,66 @@
+---
+title: Chroma
+description: Chroma vector database integration for Genkit RAG workflows.
+---
+
+The Chroma plugin registers retrievers and indexers backed by a [Chroma](https://www.trychroma.com/) server (v2 REST API) for Retrieval-Augmented Generation.
+
+## Installation
+
+```xml
+<dependency>
+    <groupId>com.google.genkit</groupId>
+    <artifactId>genkit-plugin-chroma</artifactId>
+    <version>1.0.0-SNAPSHOT</version>
+</dependency>
+```
+
+## Requirements
+
+- A running Chroma server (`docker run -p 8000:8000 chromadb/chroma`)
+- Java 21+
+- An embedder (e.g. from the Google GenAI plugin)
+
+## Usage
+
+`ChromaPlugin` registers a retriever and indexer named `chroma/<collectionName>` for each configured collection.
+
+```java
+import com.google.genkit.plugins.chroma.ChromaCollectionConfig;
+import com.google.genkit.plugins.chroma.ChromaPlugin;
+
+Genkit genkit = Genkit.builder()
+    .plugin(GoogleGenAIPlugin.create(apiKey))
+    .plugin(
+        ChromaPlugin.builder()
+            .url("http://localhost:8000")   // default
+            .addCollection(
+                ChromaCollectionConfig.builder()
+                    .collectionName("films")
+                    .embedderName("googleai/gemini-embedding-001")
+                    .distance(ChromaCollectionConfig.Distance.COSINE)
+                    .createCollectionIfNotExists(true) // default
+                    .build())
+            .build())
+    .build();
+
+// Index and retrieve
+genkit.index("chroma/films", documents);
+List<Document> results = genkit.retrieve("chroma/films", "a Christopher Nolan sci-fi film");
+```
+
+## Configuration
+
+Tune per-collection settings with `ChromaCollectionConfig`:
+
+- `collectionName` — the Chroma collection name (required)
+- `embedderName` — the embedder used to vectorize documents and queries (required)
+- `distance` — `COSINE` (default), `L2`, or `INNER_PRODUCT`
+- `createCollectionIfNotExists` — auto-create the collection (default `true`)
+- `addAdditionalMetadata(key, value)` — metadata merged into every indexed document
+
+Plugin-level `tenant` and `database` default to `default_tenant` / `default_database`.
+
+## Sample
+
+See the [chroma sample](https://github.com/genkit-ai/genkit-java/tree/main/samples/chroma).
diff --git a/docs/src/content/docs/plugins/cohere.md b/docs/src/content/docs/plugins/cohere.md
@@ -37,14 +37,29 @@ ModelResponse response = genkit.generate(
 
 ## Available models
 
+- `cohere/command-a-plus-05-2026` — flagship (Mixture-of-Experts, text + vision)
+- `cohere/command-a-reasoning-08-2025`
+- `cohere/command-a-vision-07-2025`
 - `cohere/command-a-03-2025`
 - `cohere/command-r7b-12-2024`
 - `cohere/command-r-08-2024`
 - `cohere/command-r-plus-08-2024`
 
+## Embeddings
+
+Cohere embedding models are available through the OpenAI-compatible endpoint:
+
+- `cohere/embed-v4.0`
+- `cohere/embed-multilingual-v3.0`
+- `cohere/embed-english-v3.0`
+
+```java
+EmbedResponse response = genkit.embed("cohere/embed-v4.0", documents);
+```
+
 ## Features
 
-- Text generation, tool calling, RAG support, SSE streaming
+- Text generation, tool calling, RAG support, SSE streaming, embeddings
 
 ## Sample