Incorporate AI into your phpBB board and get ready for an exciting experience.
Currently supported Midjourney, ChatGPT and DALL-E (OpenAI), Gemini and Gemini Vision (Google), Claude (Anthropic), Stable Diffusion v2/v3 (Stability AI), Pika (Pika.art), PixVerse (PixVerse.ai), FaceSwap (Picsi.Ai).
- Examples
- Requirements
- Important notes
- Installation
- BBCode
mp4
tag - Discord CDN smart proxy setup
- Midjourney setup
- FaceSwap setup
- Pika setup
- PixVerse setup
- PixVerse_MemeFace setup
- ChatGPT setup
- ChatGPT advanced setup
- Claude setup
- Gemini setup
- Gemini Vision setup
- Chat bots can share conversation history
- DALL-E setup
- DALL-E advanced features
- Stable Diffusion v3 setup
- Stable Diffusion v2 setup
- Troubleshooting
- Support and suggestions
- Changelog
- License
- Midjourney
- FaceSwap by Picsi.Ai
- Pika β’ AI text/text+image to video by Pika.art
- PixVerse β’ AI text/text+image to video by PixVerse.ai
- PixVerse_MemeFace β’ AI text+image to video by PixVerse.ai
- ChatGPT
- ChatGPT, custom prompt
- Claude
- Gemini
- Gemini Vision
- DALL-E
- Stable Diffusion v3 by Stability AI
- Stable Diffusion v2 by Stability AI
- php >=7.4
- phpbb >= 3.2
-
Installing of Simple mentions phpBB extension strongly suggested.
@mention feature makes it really easy to talk to AI bots and other board users.
π I created a small update for version 2.0 to support notifications when editing an already submitted post. Simply replace your/ext/paul999/mention/event/main_listener.php
with the provided main_listener.php to enable this feature. -
If you are planning to use image generation AI (eg DALL-E or Stable Diffusion) make sure to adjust attachment settings to support large images and verify that
webp
image extension configured.Go to
ACP
>General
>Attachment settings
and adjustTotal attachment quota
,Maximum file size
andMaximum file size messaging
:Go to
ACP
>Posting
>Manage attachment extensions
, look forwebp
, add it if missing:
Above does not apply to Discord AI bots (eg Midjourney), as all generated images are actually stored on your Discord account and served via the Discord CDN smart proxy.
-
If you have extensions installed that require users to log in, such as Login Required you will need to whitelist
/ailabs/*
and/app.php/ailabs/*
since AI Labs extension uses callbacks. -
Adjust PHP configuration to allow longer script execution. ChatGPT API responses may take up to 90 seconds to respond in certain cases. If you have the default settings, your SQL connection will be closed after 30 seconds, preventing the extension from functioning properly.
Suggested values forphp.ini
:max_execution_time = 180
max_input_time = 90
Download https://github.com/privet-fun/phpbb_ailabs and copy /privet/ailabs
to phppp/ext
folder:
IMPORTANT
If you have a previous version of this extension installed, you will need to disable it and then enable it again after the new version has been copied over.
Do not forget to purge the cache after the new version has been installed.
Go to ACP
> Customise
> Manage extensions
and enable the AI Labs
extension.
Finally go to ACP
> Extensions
> AI Labs
> Settings
and add desired AI configurations:
The following Discord AI bots are generating mp4
videos and will require adding the [mp4]
BBCode tag:
To add mp4
BBCode tag, go to ACP
> POSTING
> BBCodes
and add mp4
tag as shown below:
BBCode usage:
[mp4]{URL}[/mp4]
HTML replacement:
<video src="{URL}" style="width:100%;max-width:640px" controls>Your browser does not support the video tag.</video>
You can adjust above max-width:NNNpx
to desired value.
Help line:
[mp4]http://example.com/video.mp4[/mp4]
Discord CDN attachments have an expiration time, as seen in this original announcement screenshot. To allow users to access media generated by Discord AI bots you need to set up a Discord CDN smart proxy.
Currently following Discord AI bots supported by this plugin via API provided by useapi.net:
You will need at least one Discord token(s) from any of above Discord AI bots. In fact, you can use any Discord account to obtain a token since Discord does not check or enforce any security validations. All you need is an active Discord account and a Discord token for that account - see how to obtain Discord token. You may configure multiple Discord tokens to spread the load and avoid Discord 429 response codes. However, this is a theoretical safety measure as I have yet to encounter a 429 response for this particular use case, even with a single Discord account. It appears that Discord does not throttle attachment refresh API calls at this point.
Discord attachments use one of the following formats:
- https://cdn.discordapp.com/attachments/`channel`/`message`/`attachment_file_name`.`file_ext`?ex=`expire_date`&is=`issued_on`&hm=`check_sum`
- https://media.discordapp.net/attachments/`channel`/`message`/`attachment_file_name`.`file_ext`?ex=`expire_date`&is=`issued_on`&hm=`check_sum`
Where expire_date
and issued_on
values are unix/epoch hex timestamps, here's convertor to human-readable format.
All Discord AI bots provided by this extension will convert the original attachment links as follows:
your_website_root_url
/ailabs/discord_cdn/cdn.discordapp.com/attachments/channel
/message
/attachment_file_name
?ext=file_ext
&ex=expire_date
&is=issued_on
&hm=check_sum
your_website_root_url
/ailabs/discord_cdn/media.discordapp.net/attachments/channel
/message
/attachment_file_name
?ext=file_ext
&ex=expire_date
&is=issued_on
&hm=check_sum
Example of original attachment link (this link will likely expire by the time you read this) and adjusted Discord CDN smart proxy link (this link will be auto-refreshed by the Discord CDN smart proxy running at https://privet.fun).
Configuration steps:
- Create a new board user who will act as an AI placeholder bot, for our example we will use the user
DiscordCDN
. Make sure this user account is NOT active as we only need this account as a placeholder. - Go to
ACP
>Extensions
>AI Labs
>Settings
and add new configuration, selectdiscord_cdn
from AI dropdown:
- Use
Load default configuration/template
to get defaults.
Replace Configuration JSONrequired_discord_token_a
with Discord token used by any of configured Discord AI bots or by any other active Discord token.
Optionally you can add more Discord tokens, seeoptional_discord_token_b
andoptional_discord_token_c
or delete bothoptional_discord_token_...
if you're not planing to use them. - Update the
channels
array to include Discord channels from which you want to proxy Discord CDN attachment links. These should be the same Discord channels as the ones configured for your Discord AI bots and any other channels you optionally want to add. Thechannels
array acts as a security measure to prevent third parties from using your proxy for their own needs. - Specify
ALL
for Quote or mention, this is not really used.
- Use
To verify that your setup is working ask the Midjourney or any other configured Discord AI bot to generate something.
Alternatively you can take a valid Discord CDN link and paste it to you browser as your_website_root_url
/ailabs/discord_cdn/cdn.discordapp.com/attachments/channel
/message
/attachment_file_name
?ext=file_ext
, if Discord CDN smart proxy is configured properly it will refresh new then link and redirect your browser to that link.
If you had a version of this extension installed prior to v1.0.10, you likely already have expired Discord CDN links. Below are handy SQL statements to update your MySQL/MariaDB database to use the Discord CDN smart proxy link format. Run them one at a time and please make sure to take a backup before you make any changes.
To get rough idea how many posts will be affected:
SELECT count(*)
FROM <you_prefix>_posts
WHERE post_text LIKE "%https://media.discordapp.net/attachments/%" OR post_text LIKE "%https://cdn.discordapp.com/attachments/%"
Update posts:
UPDATE <you_prefix>_posts
SET post_text = REGEXP_REPLACE(post_text,
'https://media\\.discordapp\\.net/attachments/([0-9]+)/([0-9]+)/([^.]+)\\.([A-Za-z0-9]+)(\\?ex=)?',
'<your_website_root_url>/ailabs/discord_cdn/media.discordapp.net/attachments/\\1/\\2/\\3?ext=\\4&ex=')
WHERE post_text LIKE "%https://media.discordapp.net/attachments/%"
UPDATE <you_prefix>_posts
SET post_text = REGEXP_REPLACE(post_text,
'https://cdn\\.discordapp\\.com/attachments/([0-9]+)/([0-9]+)/([^.]+)\\.([A-Za-z0-9]+)(\\?ex=)?',
'<your_website_root_url>/ailabs/discord_cdn/cdn.discordapp.com/attachments/\\1/\\2/\\3?ext=\\4&ex=')
WHERE post_text LIKE "%https://cdn.discordapp.com/attachments/%"
-
βοΈ Midjourney tutorial and examples.
-
You'll need Midjourney Discord and useapi.net accounts with active useapi.net subscription.
-
Create new board user who will act as AI bot, for our example we will use user
Midjourney
.
Make sure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add new configuration, selectmidjourney
from AI dropdown:
- Use
Load default configuration/template
to get defaults.
Replace Configuration JSONapi-key
,discord
,server
andchannel
with your values. If you explicitly configured your Midjourney useapi.net account you do not need to specify thediscord
,server
andchannel
values and can delete them from the configuration. - Select forums where you want
Midjourney
AI user to reply to new posts and/or to quoted and @mention (if you are using Simple mentions extension) posts.
- Use
-
Save changes, navigate to forum configured above and create new post (if you configured
Reply on a post
) or quote/@mentionMidjourney
user:
-
You MUST configure the Discord CDN smart proxy so that the generated images become visible on your board.
-
βοΈ FaceSwap tutorial and examples.
-
You'll need InsightFaceSwap Discord and useapi.net accounts with active useapi.net subscription.
InsightFaceSwap Discord bot offers free and paid subscription models.
If you're planning to use the free version, you may want to consider adding a second account since the free version comes with only 50 credits per day. Based on your usage, you may run out of free credits. Refer to useapi.net instructions on how to set up multiple InsightFaceSwap accounts. -
Create new board user who will act as AI bot, for our example we will use user
FaceSwap
.
Make sure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add new configuration, selectfaceswap
from AI dropdown:
-
Use
Load default configuration/template
to get defaults.
Replace Configuration JSONapi-key
andchannel
with your values. If you explicitly configured your InsightFaceSwap useapi.net account you do not need to specifychannel
value and can delete it from the configuration. -
Select forums where you want
FaceSwap
AI user to reply to new posts and/or to quoted and @mention (if you are using Simple mentions extension) posts.
-
-
Save the changes, navigate to the forum configured above, and create a new post (if you configured
Reply on a post
) or quote/@mention theFaceSwap
user to verify that it is working as expected. Refer to the troubleshooting section if you encounter any issues. -
You MUST configure the Discord CDN smart proxy so that the generated images become visible on your board.
-
βοΈ Pika tutorial and examples.
-
You'll need Pika Discord and useapi.net accounts with active useapi.net subscription.
Pika Discord bot is currently free to use and does not require any additional subscription. -
Create new board user who will act as AI bot, for our example we will use user
Pika
.
Make sure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add new configuration, selectpika
from AI dropdown:
-
Use
Load default configuration/template
to get defaults.
Replace Configuration JSONapi-key
,discord
andchannel
with your values. If you explicitly configured your Pika useapi.net account you do not need to specify thediscord
andchannel
values and can delete them from the configuration. -
Select forums where you want
Pika
AI user to reply to new posts and/or to quoted and @mention (if you are using Simple mentions extension) posts.
-
-
Save changes, navigate to forum configured above and create new post (if you configured
Reply on a post
) or quote/@mentionPika
user:
-
Refer to this post to learn more about the currently supported Pika bot functionality.
-
You MUST configure the Discord CDN smart proxy so that the generated mp4 media become visible on your board.
-
You MUST configure the BBCode
mp4
tag so that the generated mp4 media become viewable on your board.
-
βοΈ PixVerse tutorial and examples.
-
You'll need PixVerse Discord and useapi.net accounts with active useapi.net subscription.
PixVerse Discord bot is currently free to use and does not require any additional subscription. -
Create new board user who will act as AI bot, for our example we will use user
PixVerse
.
Make sure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add new configuration, selectpixverse
from AI dropdown:
-
Use
Load default configuration/template
to get defaults.
Replace Configuration JSONapi-key
,discord
,server
andchannel
with your values.
If you explicitly configured your PixVerse useapi.net account you do not need to specifydiscord
,server
andchannel
values and can delete them from the configuration.
π You can replace theurl_create
field value withhttps://api.useapi.net/v1/pixverse/create_single
if you wish to generate a single video per request. If you choose to do so, it is safe to increase themaxJobs
value to5
or perhaps even7
. -
Select forums where you want
PixVerse
AI user to reply to new posts and/or to quoted and @mention (if you are using Simple mentions extension) posts.
-
-
Save the changes, navigate to the forum configured above, and create a new post (if you configured
Reply on a post
) or quote/@mention thePixVerse
user to verify that it is working as expected. Refer to the troubleshooting section if you encounter any issues. -
You MUST configure the Discord CDN smart proxy so that the generated mp4 media become visible on your board.
-
You MUST configure the BBCode
mp4
tag so that the generated mp4 media become viewable on your board.
-
βοΈ PixVerse_MemeFace tutorial and examples.
-
You'll need PixVerse Discord and useapi.net accounts with active useapi.net subscription.
PixVerse Discord bot is currently free to use and does not require any additional subscription. -
Create new board user who will act as AI bot, for our example we will use user
PixVerse_MemeFace
.
Make sure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add new configuration, selectpixverse_meme_face
from AI dropdown:
-
Use
Load default configuration/template
to get defaults.
Replace Configuration JSONapi-key
,discord
,server
andchannel
with your values.
If you explicitly configured your PixVerse useapi.net account you do not need to specifydiscord
,server
andchannel
values and can delete them from the configuration. -
Select forums where you want
PixVerse_MemeFace
AI user to reply to new posts and/or to quoted and @mention (if you are using Simple mentions extension) posts.
-
-
Save the changes, navigate to the forum configured above, and create a new post (if you configured
Reply on a post
) or quote/@mention thePixVerse_MemeFace
user to verify that it is working as expected. Refer to the troubleshooting section if you encounter any issues. -
You MUST configure the Discord CDN smart proxy so that the generated mp4 media become visible on your board.
-
You MUST configure the BBCode
mp4
tag so that the generated mp4 media become viewable on your board.
-
βοΈ ChatGPT tutorial and examples.
-
You will need OpenAI account, sign up at https://platform.openai.com/.
To obtain API key go to https://platform.openai.com/account/api-keys, click onCreate new secret key
, copy and save in a safe place generated API key.
Open AI key starts withsk-
and look something like thissk-rb5yW9j6Nm2kP3Fhe7CPzT1QczwDZ5LvnlBfYU2EoqyX1dWs
. -
Create new board user who will act as AI bot, for our example we will use user
ChatGPT
.
Make sure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add new configuration, selectchatgpt
from AI dropdown:
- Use
Load default configuration/template
to get defaults.
Replace Configuration JSONapi-key
with your Open AI key. - Select forums where you want
ChatGPT
AI user to reply to new posts and/or to quoted and @mention (if you are using Simple mentions extension) posts.
- Use
-
Save changes, navigate to forum configured above and create new post (if you configured
Reply on a post
) or quote/@mentionChatGPT
user:
-
Fine-tuning can be done by adjusting following OpenAI API chat parameters https://platform.openai.com/docs/api-reference/chat
model
, defaultgpt-4o
, full list of models available at https://platform.openai.com/docs/modelstemperature
,top_p
,frequency_penalty
andpresence_penalty
- see https://platform.openai.com/docs/api-reference/chat/create
-
Additional setting used by ChatGPT AI
max_tokens
, default 4096, define size reserved for AI reply when quoted.prefix
, default empty, can be used to prompt model, see ChatGPT advanced setup for details.
π You can hint ChatGPT to format its responses using phpBB bbcodes markup by specifying"prefix": "Use phpBB bbcode formatting when answering, unless explicitly specified otherwise."
and"prefix_tokens": 21
.prefix_tokens
, default 0, see ChatGPT advanced setup for details.max_quote_length
, if provided, the quoted response text will be truncated to the number of words defined by the max_quote_length value. Set it to 0 to remove all quoted text entirely.
You can setup ChatGPT to pretend it is somebody else using param prefix
with custom prompt (aka system prompt).
Let's create new board user Bender
and configure it same as we did in ChatGPT setup .
We want use prefix
and prefix_tokens
params to fine-tune ChatGPT AI behavior so our AI bot Bender
will provide responses like this, mostly staying in a character.
To determine what number should be placed in prefix_tokens
let's ask our freshly created AI bot Bender
question which we want to use for prefix
.
For example below we will use for prefix
following system prompt Pretend your are Bender from Futurma
Once bot replied click on log icon, and note value of Request tokens
.
Finally go back to Bender
AI bot configuration and update params prefix
and prefix_tokens
-
βοΈ Claude tutorial and examples.
-
Please follow the Anthropic instructions to create and activate a Claude API key.
Note the Claude API key you create, you will need it later to set up the Claude bot. -
Create a new board user who will act as the AI bot. For our example, we will use the user
Claude
.
Ensure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add a new configuration, selectingclaude
from the AI dropdown:
- Use
Load default configuration/template
to load the defaults.
Replace<API-KEY>
in the Configuration JSON with your Claude API key. - Select the forums where you want the
Claude
AI user to reply to new posts and/or to quoted and @mention posts (if you are using the Simple Mentions extension).
- Use
-
Save the changes, navigate to the forum configured above, and create a new post (if you configured
Reply on a post
) or quote/@mention theClaude
user to verify that it is working as expected. Refer to the troubleshooting section if you encounter any issues. -
Fine-tuning can be achieved by adjusting the following Claude API configuration parameters:
model
can be found here: https://docs.anthropic.com/claude/docs/models-overview#model-recommendations.temperature
,max_tokens
,system
can be found here: https://docs.anthropic.com/claude/reference/messages_post. Parametersystem
is a way of providing context and instructions to Claude, such as specifying a particular goal or role, see guide to system prompts. If specified you will need to add number of tokens used by system prompt tosystem_token
value to ensure correct token count. You can follow instructions for ChatGPT advanced setup to calculatesystem_token
value.
π You can hint Claude to format its responses using phpBB bbcodes markup by specifying"system": "Use phpBB bbcode formatting when answering, unless explicitly specified otherwise."
and"system_tokens": 23
.
Users can overridetemperature
parameter by providing a hint in the message using the--temperature value
notation, e.g.--temperature 0
or--temperature 0.5
-
Additional settings:
max_quote_length
, if provided, the quoted response text will be truncated to the number of words defined by themax_quote_length
value. Set it to 0 to remove all quoted text entirely.
-
βοΈ Gemini tutorial and examples.
-
Please follow the Google instructions to create and activate a Gemini API key in Google AI Studio.
Note the Gemini API key you create, you will need it later to set up the Gemini and Gemini Vision bots. -
Create a new board user who will act as the AI bot, for our example, we will use the user
Gemini
.
Ensure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add a new configuration, selectinggemini
from the AI dropdown:
- Use
Load default configuration/template
to load the defaults.
Replace<API-KEY>
in the Configuration JSON with your Gemini API key.
π You can hint Gemini to format its responses using phpBB bbcodes markup by specifying"prefix": "Use phpBB bbcode formatting when answering, unless explicitly specified otherwise."
. - Select the forums where you want the
Gemini
AI user to reply to new posts and/or to quoted and @mention posts (if you are using the Simple Mentions extension).
- Use
-
Save the changes, navigate to the forum configured above, and create a new post (if you configured
Reply on a post
) or quote/@mention theGemini
user to verify that it is working as expected. Refer to the troubleshooting section if you encounter any issues. -
Fine-tuning can be achieved by adjusting the following Gemini API configuration parameters:
model
can be found here: https://ai.google.dev/models/gemini#model-versions,model
is part ofurl_generateContent
andurl_countTokens
.temperature
,topK
,topP
can be found here: https://ai.google.dev/docs/concepts#model_parameters, these should be placed in thegeneration_config
node. Users can override the above parameters by providing a hint in the message using the--param value
notation, where--param
is case-insensitive. E.g.--temperature 0
or--temperature 0.5 --topk 1 --topp 0.8
-
Additional settings used by the Gemini API:
max_tokens
, default 30720, this is the maximum size of the entire conversation.prefix
, default is empty, it can be used to prompt the model.max_quote_length
, if provided, the quoted response text will be truncated to the number of words defined by themax_quote_length
value. Set it to 0 to remove all quoted text entirely.
For an examples of how to use Gemini bot please refer to Gemini.
βοΈ Gemini Vision tutorial and examples.
The setup for Gemini Vision follows the same steps as the above-mentioned Gemini bot. You will need to create a separate board user, e.g. GeminiVision
and select gemini_vision
from the AI dropdown.
The Gemini Vision bot does not support conversations, you will need to provide a prompt along with an image every time. You can attach an image to the post or provide an image URL directly in the prompt. For an examples of how to use Gemini Vision bot please refer to Gemini Vision.
AI chat bots (ChatGPT, Gemini, and Claude) can now access each other's conversation history and context (starting from version 1.0.10 of this extension).
You can start chatting with one AI chat bot and later on in the conversation tag another bot(s).
Tagged bots will automatically inherit the entire conversation history and context.
Please see example.
βοΈ DALL-E tutorial and examples.
Setup mostly the same as for ChatGPT above:
Refer to https://platform.openai.com/docs/api-reference/images/create to learn more about n
and size
parameters.
Examples
-
To generate an image of the desired size, you can specify one of the following sizes anywhere within the prompt, example:
- 1024x1024
- 512x512
- 256x256
-
To create variations of the image simply post image url to the prompt, example
-
βοΈ Stable Diffusion v3 tutorial and examples.
-
You will need Stability AI account, follow official instructions to create account and obtain API key.
-
Create a new board user who will act as the AI bot, for our example, we will use the user
Stable Diffusion v3
.
Ensure this user account is activated and fully functional. -
Go to
ACP
>Extensions
>AI Labs
>Settings
and add a new configuration, selectingstablediffusion_v3
from the AI dropdown:
- Use
Load default configuration/template
to load the defaults.
Replace<API-KEY>
in the Configuration JSON with your Stability AI account API key. - Select the forums where you want the
Stable Diffusion v3
AI user to reply to new posts and/or to quoted and @mention posts (if you are using the Simple Mentions extension).
- Use
-
Save the changes, navigate to the forum configured above, and create a new post (if you configured
Reply on a post
) or quote/@mention theStable Diffusion v3
user to verify that it is working as expected. Refer to the troubleshooting section if you encounter any issues.
-
This section covers original Stable Diffusion v2 setup.
-
You will need Stability AI account, follow official instructions to create account and obtain API key.
-
Create new board user, let's say
Stable Diffusion
and create configuration:
Examples -
Refer to https://api.stability.ai/docs#tag/v1generation/operation/textToImage to learn more about configuration JSON parameters.
-
AI Labs extension maintains internal logs, you should have admin or moderator rights to see log icon
You can see entire AI communication history in the log:
If Log entry is empty it usually means that/ailabs/*
or/app.php/ailabs/*
routes blocked by one of phpBB extensions (eg Login Required) and you will need to add/ailabs/*
or/app.php/ailabs/*
to extension whitelist.
You can examine Logresponse
(JSON) to see details for AI response.
Please feel free to post your questions or concerns at https://github.com/privet-fun/phpbb_ailabs/issues. -
When setting up your bot, you will be able to test the bot URL by referring to the
Bot URL (test)
link below
If you do not see the bot responseProcessing job 0
, you will need to investigate what is preventing access to that URL, your web server logs will be good place to start. -
You can enable cURL communication logging by adding the
"debug": true
parameter to your bot configuration. The AI Labs extension uses cURL to communicate with AI APIs. By enabling logging, you should be able to see the entire data exchange between the extension and the AI APIs. Look for/var/www/phpbb/curl_debug.txt
(or similar) for log content.
This extension is currently being actively developed. For communication, please use https://github.com/privet-fun/phpbb_ailabs/issues.
-
2.0.0 June 30, 2024
- Support for four new Discord AI bots added:
- Link to API help page along with the API command will be provided within the reply (where applicable) for all Discord AI bots.
- All default AI bot templates have been updated with the most recent configuration values, including model names and other parameters.
- Added citation links are optionally provided by Gemini 1.5+. Notice the example URL links at the bottom of the reply.
- Field sizes of the request and response fields in the log table have been increased to accommodate longer content.
- Major codebase refactoring to unify logic for Discord AI bots supported via useapi.net and to handle attachments and images imported from URLs.
- Minor bug fixes.
IMPORTANT
When updating to v 2.0.0 from older versions you will need to updated your existing configuration(s). After then new version is installed go thru all configured AI bots andLoad default configuration/template
for each, make sure to save your API keys and Discord configuration (discord token, server and channel values where applicable) and place them back into updated configuration. -
1.0.10 May 5, 2024
- Added Discord CDN smart proxy setup
- Added Italian transaction by Lord Phobos
-
1.0.9 April 17, 2024
- Added support for Pika by Pika.art AI text/text+image to video bot
- Added support for Claude by Anthropic
- AI chat bots (ChatGPT, Gemini and Claude) can now share each other's conversation history and context example
- The troubleshooting features have been greatly extended
- You can edit the original conversation after it has been posted and add more
@mention
AI bot tags if you missed them example - Created small update for Simple mentions phpBB extension version 2.0 to support notifications when editing an already submitted post. Refer to Important notes for details
-
1.0.8 March 10, 2024
- Added support for Gemini and Gemini Vision by Google
- Added support for Simple mentions phpBB extension version 2.x
-
1.0.7 December 26, 2023
- Updated the Midjourney Bot to support the v2 API from https://useapi.net
Make sure to update your Midjourney bot configuration:"url_imagine": "https://api.useapi.net/v2/jobs/imagine", "url_button": "https://api.useapi.net/v2/jobs/button",
- All messages and warnings from the Midjourney Bot will now be relayed back
- Custom Zoom support added
- Added support for Midjourney v6, including quoted text and new buttons
- Updated the Midjourney Bot to support the v2 API from https://useapi.net
-
1.0.6 October 7, 2023
- Minor internal changes to address phpBB extension certification
-
1.0.5 October 1, 2023
- Midjourney support added
max_quote_length
option added for ChatGPT
-
1.0.4 June 4, 2023
- Troubleshooting section added
- Added configuration for reply in topics
- Fixed links generation for cases where cookies disabled
- AI Labs internal controllers (
/ailabs/*
) will attempt to establish session to deal with phpBB extensions like Login Required - Better descriptions added to help with setup
- Minor bugfixes
-
1.0.3 June 1, 2023
- bumped php requirements to >= 7.4
- Comma removed, reported by Vlad__
-
1.0.2 June 1, 2023
- Only apply
utf8_encode_ucr
if present, reported by Vlad__
This will allow phpBB 3.2.1 support without any modifications. - Removed
...
andarray
to support php 7.x, reported by Vlad__ - Added missing
reply
processing for chatgpt controller, reported by Vlad__ - Added board prefix to all links, reported by Miri4ever
- Only apply
-
1.0.1 May 29, 2023
- Fixed issues reported by Miri4ever
- Removed all MySQL specific SQL, going forward extension should be SQL server agnostic
- Better language management
- Minor code cleanup
-
1.0.0 May 28, 2023
- Public release