Docs for deploying SeaTable AI

docs/configuration/roles-and-permissions.md

@@ -45,6 +45,7 @@ The following quotas are supported in user roles:
 | scripts_running_limit          | 2.3              | Total number of _Python_ scripts run within a month: 100 means 100 script runs per month; -1 means unlimited script runs                                         | The script run counter is reset at the beginning of every month.  |
 | snapshot_days                  | 2.1              | Retention period for snapshots in days: 180 means a storage period of 180 days; no value means an unlimited retention period                                     | Snapshots older than the retention period are automatically removed.   |
 | share_limit                    |                  | Max number of users a base can be shared with: 100 means a base can be shared with 100 users             |    |
+| ai_credit_per_user                  | 6.0              |   The maximum AI quota allowed per user per month (i.e., the maximum amount of tokens that can be used in a single month, converted into an amount. In team mode, the total quota within the team will be shared). `-1` means unlimited quota.                                   |    |
 
 
 ### Standard User Roles

docs/installation/advanced/seatable-ai-standalone.md

@@ -0,0 +1,86 @@
+# Standalone Deployment of SeaTable AI
+
+This guide describes the standalone deployment of `seatable-ai` on a dedicated server or virtual machine.
+
+## Prerequisites
+
+- You have successfully installed [Docker and Docker-Compose](../basic-setup.md#install-docker-and-docker-compose-plugin)
+- You have [downloaded the latest `.yml` files](../basic-setup.md#1-create-basic-structure) from the `seatable-release` GitHub repository
+- The hosts destined to run `seatable-ai` and other SeaTable components are attached to the same private network
+
+## SeaTable AI Configuration
+
+The following section outlines an `.env` file with the settings needed to run `seatable-ai`.
+These changes should be made inside `/opt/seatable-compose/.env`:
+
+```ini
+COMPOSE_FILE='seatable-ai-standalone.yml'
+COMPOSE_PATH_SEPARATOR=','
+
+# system settings
+TIME_ZONE='Europe/Berlin'
+
+# database
+MARIADB_HOST=
+MARIADB_PORT=3306
+MARIADB_PASSWORD=
+
+# redis
+REDIS_HOST=
+REDIS_PORT=6379
+REDIS_PASSWORD=
+
+# This private key must have the same value as the JWT_PRIVATE_KEY variable on other SeaTable nodes
+JWT_PRIVATE_KEY=
+
+# Public URL of your SeaTable server
+SEATABLE_SERVER_URL=https://seatable.your-domain.com
+
+# Cluster-internal URL of dtable-server
+INNER_DTABLE_SERVER_URL=http://dtable-server:5000
+
+# Cluster-internal URL of dtable-db
+INNER_DTABLE_DB_URL=http://dtable-db:7777
+
+# LLM
+SEATABLE_AI_LLM_TYPE=openai
+SEATABLE_AI_LLM_URL=
+SEATABLE_AI_LLM_KEY=...
+SEATABLE_AI_LLM_MODEL=gpt-4o-mini # recommended
+```
+
+!!! warning
+    - In case you are not using password authentication for Redis, you should not specify a value after the equals sign (`=`) for the `REDIS_PASSWORD` variable.
+      Specifying an empty string (e.g. `REDIS_PASSWORD=""`) will cause problems.
+
+    - By default, the ports of `dtable-server` (5000) and `dtable-db` (7777) are not exposed to the host. This requires a manual change inside the `.yml` file.
+
+### LLM Provider Configuration
+
+Please refer to the documentation on [configuring your LLM provider of choice](../components/seatable-ai.md#llm-provider-configuration).
+These configuration details do not change depending on the deployment topology of `seatable-server` and `seatable-ai`.
+
+### Start SeaTable AI
+
+You can now start SeaTable AI by running the following command inside your terminal:
+
+```bash
+cd /opt/seatable-compose
+docker compose up -d
+```
+
+## Configuration of SeaTable Server
+
+Since `seatable-ai` is now running on a separate host or virtual machine, the following configuration changes must be made inside the `.env` file on the host running the `seatable-server` container:
+
+```ini
+ENABLE_SEATABLE_AI=true
+SEATABLE_AI_SERVER_URL='http://seatable-ai.example.com:8888'
+```
+
+Restart the `seatable-server` service and test your SeaTable AI:
+
+```bash
+cd /opt/seatable-compose
+docker compose up -d
+```

docs/installation/advanced/seatable-ai-token-pricing.md

@@ -0,0 +1,32 @@
+# AI Token Pricing
+
+## AI Credits
+
+AI credits serve as an internal unit of currency for measuring AI-related usage within SeaTable.
+They are directly linked to the number of tokens consumed by using AI-based features according to the [configured price](#pricing-configuration) of each AI model.
+
+SeaTable supports role-based AI credit limits by configuring the `ai_credit_per_user` option on a user role.
+Please refer to the documentation on [user quotas](../../configuration/roles-and-permissions.md#user-quotas) for more details.
+
+!!! note "`ai_credit_per_user` for organization users"
+    AI credits are shared across all users inside a SeaTable organization. The total number of credits can be calculated by multiplying the value of `ai_credit_per_user` by the number of team users.
+
+    **Example:** Setting `ai_credit_per_user` to `2` will allow a team with 10 members to have 20 AI credits in total.
+
+## Pricing Configuration
+
+In order to accurately track the number of AI credits used by users and organizations, you must configure token pricing inside `/opt/seatable-server/seatable/conf/dtable_web_settings.py`.
+This can be achieved by configuring the `AI_PRICES` variable, which is a dictionary that maps model identifiers (e.g `gpt-4o-mini`) to token pricing **per thousand tokens**:
+
+```py
+AI_PRICES = {
+    "gpt-4o-mini": {
+        "input_tokens_1k": 0.01827, # price / 1000 tokens
+        "output_tokens_1k": 0.07309 # price / 1000 tokens
+    },
+}
+```
+
+!!! warning "Model Identifiers"
+    The dictionary key must match **the exact value** of the chosen AI Model, which is configured through the `SEATABLE_AI_LLM_MODEL` variable inside your `.env` file.
+    In case of a mismatch, AI usage will not count towards any configured credit limits!

docs/installation/components/seatable-ai.md

@@ -0,0 +1,115 @@
+# SeaTable AI Integration
+
+<!-- md:version 6.0 -->
+
+SeaTable AI is a SeaTable extension that integrates AI functionality into SeaTable.
+Deploying SeaTable AI allows users to execute AI-based automation steps within SeaTable.
+
+At the time of writing, the following types of automation steps are supported:
+
+- **Summarize**
+- **Classify**
+- **OCR** (Optical character recognition)
+- **Extract**
+- **Custom** for individual use cases
+
+## Deployment
+
+!!! note "SeaTable AI requires SeaTable 6.0"
+
+The easiest way to deploy SeaTable AI is to deploy it on the same host as SeaTable Server. A standalone deployment (on a separate host or virtual machine) is explained [here](../advanced/seatable-ai-standalone.md).
+
+### Amend the .env file
+
+To install SeaTable AI, include `seatable-ai.yml` in the `COMPOSE_FILE` variable within your `.env` file. This instructs Docker-Compose to include the `seatable-ai` service.
+
+Simply copy and paste (:material-content-copy:) the following code into your command line:
+
+```bash
+sed -i "s/COMPOSE_FILE='\(.*\)'/COMPOSE_FILE='\1,seatable-ai.yml'/" /opt/seatable-compose/.env
+```
+
+Then add SeaTable AI server configurations in `.env`:
+
+```ini
+ENABLE_SEATABLE_AI=true
+SEATABLE_AI_SERVER_URL=http://seatable-ai:8888
+```
+
+#### LLM Provider Configuration
+
+SeaTable AI will use AI functions in conjunction with a Large Language Model (LLM) service.
+
+!!! note "Supported LLM Providers"
+
+    SeaTable AI supports a wide variety of LLM providers through [LiteLLM](https://docs.litellm.ai/docs) as well as any LLM services with OpenAI-compatible endpoints. Please refer to [LiteLLM's documentation](https://docs.litellm.ai/docs/providers) in case you run into issues while trying to use a specific provider.
+
+!!! note "Model Selection"
+
+    In order to ensure the efficient use of SeaTable AI features, you need to select a **large, multimodal model**.
+    This requires the chosen model to support image input and recognition (e.g. for running OCR as part of automations).
+
+The following section showcases the required configuration settings for the most popular hosted LLM services.
+These must be configured inside your `.env` file:
+
+<a id="llm-configuration"></a>
+=== "OpenAI"
+    ```ini
+    SEATABLE_AI_LLM_TYPE=openai
+    SEATABLE_AI_LLM_KEY=<your openai LLM access key>
+    SEATABLE_AI_LLM_MODEL=gpt-4o-mini # recommended
+    ```
+=== "Deepseek"
+    ```ini
+    SEATABLE_AI_LLM_TYPE=deepseek
+    SEATABLE_AI_LLM_KEY=<your LLM access key>
+    SEATABLE_AI_LLM_MODEL=deepseek-chat # recommended
+    ```
+=== "Azure OpenAI"
+    ```ini
+    SEATABLE_AI_LLM_TYPE=azure
+    SEATABLE_AI_LLM_URL= # your deployment url, leave blank to use default endpoint
+    SEATABLE_AI_LLM_KEY=<your API key>
+    SEATABLE_AI_LLM_MODEL=<your deployment name>
+    ```
+=== "Ollama"
+    ```ini
+    SEATABLE_AI_LLM_TYPE=ollama_chat
+    SEATABLE_AI_LLM_URL=<your LLM endpoint>
+    SEATABLE_AI_LLM_KEY=<your LLM access key>
+    SEATABLE_AI_LLM_MODEL=<your model-id>
+    ```
+=== "HuggingFace"
+    ```ini
+    SEATABLE_AI_LLM_TYPE=huggingface
+    SEATABLE_AI_LLM_URL=<your huggingface API endpoint>
+    SEATABLE_AI_LLM_KEY=<your huggingface API key>
+    SEATABLE_AI_LLM_MODEL=<model provider>/<model-id>
+    ```
+=== "Self-Hosted Proxy Server"
+    ```ini
+    SEATABLE_AI_LLM_TYPE=proxy
+    SEATABLE_AI_LLM_URL=<your proxy url>
+    SEATABLE_AI_LLM_KEY=<your proxy virtual key> # optional
+    SEATABLE_AI_LLM_MODEL=<model-id>
+    ```
+=== "Other"
+    If you are using an LLM service with ***OpenAI-compatible endpoints***, you should set `SEATABLE_AI_LLM_TYPE` to `other` or `openai`, and set other LLM configuration settings as necessary:
+
+    ```ini
+    SEATABLE_AI_LLM_TYPE=...
+    SEATABLE_AI_LLM_URL=...
+    SEATABLE_AI_LLM_KEY=...
+    SEATABLE_AI_LLM_MODEL=...
+    ```
+
+### Download SeaTable AI image and restart
+
+One more step is necessary to download the SeaTable AI image and restart the SeaTable service:
+
+```bash
+cd /opt/seatable-compose
+docker compose up -d
+```
+
+Now SeaTable AI can be used.

mkdocs.yml

@@ -151,6 +151,7 @@ nav:
       - Our deployment approach: installation/deployment-approach.md
       - Single-Node Deployment:
           - SeaTable Server: installation/basic-setup.md
+          - SeaTable AI: installation/components/seatable-ai.md
           - Python Pipeline: installation/components/python-pipeline.md
           - Whiteboard: installation/components/whiteboard.md
           - n8n: installation/components/n8n.md
@@ -178,6 +179,9 @@ nav:
               - Webserver Security: installation/advanced/webserver-security.md
               - Maintenance Mode: installation/advanced/maintenance-mode.md
               - Advanced Settings for Caddy: installation/advanced/settings-caddy.md
+          - SeaTable AI:
+              - SeaTable AI (standalone): installation/advanced/seatable-ai-standalone.md
+              - AI Token Pricing: installation/advanced/seatable-ai-token-pricing.md
           - S3 Object Storage:
               - Configuration: installation/advanced/s3.md
               - Migration: installation/advanced/s3-migration.md