v221

PokiToki can now use any OpenAI-compatible providers, such as OpenAI, OpenRouter, or Nebius.

To change the provider, set the following configuration properties:

• openai.url - AI API endpoint.
• openai.api_key - API key to use the AI.
• openai.model - Chat model name.

After upgrading to v221, you can set these options without restarting the bot using the /config command. For example:

/config openai.url https://api.studio.nebius.ai/v1
/config openai.api_key ***
/config openai.model meta-llama/Meta-Llama-3.1-70B-Instruct

⚠️ Breaking changes:

• Dropped support for the DaVinci model.
• Dropped support for the Azure API.

v210

Support for o1 models.

Configurable context window size for custom models (openai.window property, see config.example.yml).

v201

Support for gpt-4o-mini model.

v199

Support for gpt-4o model (only the text mode is supported, not images, audio or video).

v197

Ask with file

To ask a question about a document, send it as a file and write the question in the caption. The bot will read the file contents and answer. Currently only supports text content (plain text, code, data), not PDFs, images or audio. Sending multiple files is also not supported.

Other changes

• Do not try to fetch quoted links (e.g. "https://example.org" or 'https://example.org').
• Updated the list of supported OpenAI models (see below).

Supported OpenAI models:

gpt-4-turbo
gpt-4-turbo-preview
gpt-4-vision-preview
gpt-4
gpt-4-32k
gpt-3.5-turbo-1106
gpt-3.5-turbo
gpt-3.5-turbo-16k

If you are using something else, please upgrade to one of the above.

v188

GPT-4 models gpt-4-turbo-preview and gpt-4-0125-preview.

v182

• Custom chat-wide prompts.
• No message history in groups (the bot only answers direct questions or follow-ups to bot messages).

v176

Support for newer OpenAI models:

gpt-4-1106-preview
gpt-4-vision-preview
gpt-4
gpt-4-32k
gpt-3.5-turbo-1106
gpt-3.5-turbo
gpt-3.5-turbo-16k

dall-e-3

The model for the imagine command is now dall-e-3 (this is currently hardcoded and can't be changed). The default image size is 1024x1024.

v165

• Support for Azure OpenAI (#18)
• Add/remove individual list items with /config (e.g. /config telegram.usernames +alice)
• Edit conversation.depth on the fly with /config

v153

• Optional message limits for groups (e.g. "10 per hour" or "1 per minute"), see the conversation.message_limit config property.
• Enable the /imagine command for specific audiences, see the imagine.enabled config property.
• Rename the max_history_depth config property to conversation.depth.