Skip to main content

Changelog

Update your image with:

docker pull promptpanel/promptpanel:1.21

and then re-create your container.

Features:

  • Added Google + Microsoft SSO support.
  • Title tags are configurable with PROMPT_TITLE variable.
  • Updated panel ordering in sidebar.
  • Bumped requirements.

Enhancements:

  • Fixed issue with blank / deleted metadata variables for panels.
  • Updated JWT logic to emit logging more effectively.
  • Converted stop token string to an array.
  • Removed aggressive cleanup from standard models.

Update your image with:

docker pull promptpanel/promptpanel:1.20

and then re-create your container.

Features:

  • Added OpenAI GPT-4o Mini model support.
  • Support for SQLite WAL settings.

Enhancements:

  • Bumped requirements.

Update your image with:

docker pull promptpanel/promptpanel:1.19

and then re-create your container.

Features:

  • Deepgram text-to-speech (TTS) and speech-to-text (STT) support.
  • Added Perplexity models to universal manifests.
  • Added controller for cancelling messages.

Enhancements:

  • Fix for filename when uploading files with the same name.
    • Didn't show the filename with the unique suffix previously, now does..
  • Env var handling cleanup.
  • Fixed documentation links across the app.
  • Tweaks to how /commands are displayed and manifests.
  • Updated starter context for non-default models to 8192.

Update your image with:

docker pull promptpanel/promptpanel:1.18

and then re-create your container.

Features:

  • Added Sonnet to Anthropic onboarding, plugin, and Universal LLM model.
  • Added "Panel frame" mode for default template and JON manifest (to link external services) [beta: coming soon].

Enhancements:

  • Fixed Deepseek listing for "simple" model fields as well as the main ones.
  • Small timezone / linking to new website tweaks.
  • Requirements bump.

Update your image with:

docker pull promptpanel/promptpanel:1.17

and then re-create your container.

Features:

  • Deepseek support (via API) in Universal LLM model [also available via Ollama locally via: deepseek-v2 and deepseek-coder-v2].
  • Copy button for code snippets.

Enhancements:

  • Summary Plugin: Fix for streaming support / simplification.
  • Made default models more resilient for aliases.
  • Tweaks to code display.
  • Requirements bump LiteLLM.

Update your image with:

docker pull promptpanel/promptpanel:1.16

and then re-create your container.

Features

  • Added vision support to Ollama / Google Gemini models.
  • Adopted the explicit /action system across all plugins.
    • /lookup for Document Lookup (RAG) plugin.
    • /append for appending a document to all base models plugins.

Enhancements

  • Summarize plugin returns one unified message now instead of multiple.
  • Fixed environment variable issue with Ollama.
  • Fixed document appending in Completion Model - added specific document_content context which can be used in Jinja templating.
  • Added Ollama call-out for when disabled.
  • Permanent fix for GPT-4o tokenizer, removed shim.
  • Improved thread titling prompt for models.
  • Improvements to Search window design.
0:00
/0:14

Update your image with:

docker pull promptpanel/promptpanel:1.15

and then re-create your container.

Features

  • New document summarization model.
  • Added User Signup / Activation flows (optional, disabled by default) [PROMPT_USER_SIGNUP / PROMPT_USER_SIGNUP_ACTIVATE ].
  • ] Added Password Reset flow (optional, disabled by default) [PROMPT_USER_RESET_PASSWORD.
  • New explicit /actions support.
  • Citations now supported by default template (instead of custom template).
  • Added email SMTP credentials (used for activation / password reset):
    • PROMPT_SMTP_HOST
    • PROMPT_SMTP_USE_TLS
    • PROMPT_SMTP_PORT
    • PROMPT_SMTP_USER
    • PROMPT_SMTP_PASSWORD
    • PROMPT_SMTP_FROM_ADDRESS

Enhancements

  • JWT refresh logic updates.
  • Added specific blacklist for tokens (instead of full log for bloat).
  • Updated Tailwind, fixed prose.
  • Small visual fixes

Update your image with:

docker pull promptpanel/promptpanel:1.14

and then re-create your container.

Features

  • Added search fronend and updated search API endpoint (beta).

Enhancements

  • Added file status to file uploads (stopped auto-deleting after failure in plugins).
  • Added configurable upload file size max with PROMPT_MAX_FILESIZE.
  • Added refresh token to work with access token.
  • Added configurable token expiry times with PROMPT_ACCESS_TOKEN_EXP_MINUTES and PROMPT_REFRESH_TOKEN_EXP_MINUTES.
  • Invalidated JWTs on logout.
  • Fix for streaming responses when using Nginx:
  • Updated thread titling method for plugins.
  • Added Dockerfile.

Update your image with:

docker pull promptpanel/promptpanel:1.13

and then re-create your container.

Features

  • Added pagination to API, updated templates for Message pagination (stubs for Files / Threads).
  • Added search endpoint (beta) to API.

Enhancements

  • Updated completion model to calculate tokens just-in-time.
  • Visual tweaks / fixes.
  • Fixed update panel endpoint (allow all adminstrators to update panels).

Update your image with:

docker pull promptpanel/promptpanel:1.12

and then re-create your container.

Features

  • Added Mistral open, Mistral Codestral models.
  • Added Gemini Flash 1.5 to universal models.

Enhancements

  • Bumped requirements.
  • Refactored RAG plugin to be compatible with more models.
  • Breaking: RAG system rename to Document Lookup.

Update your image with:

docker pull promptpanel/promptpanel:1.11

and then re-create your container.

Enhancements

  • Modified update user to endpoint to disable user email update until SMTP confirmation flow added
  • Fix for OpenAI token model issue (temp__token_model)
  • Fix for Gemini Flash model handler (str not int issue)
  • Fix for RAG dependancies (poppler, tesseract)
  • Visual fixes

Update your image with:

docker pull promptpanel/promptpanel:1.10

and then re-create your container.

Features

  • Added Gemini Flash support
  • Added RAG documents plugin
  • Added media permissions to match panel permissions

Enhancements

  • Fixed filepath issue for thread media
  • Fixes GPT plugin for model naming (GPT-4 Turbo vs. GPT-4)
  • Temporary token fix for GPT-4o
  • Fixes for file metadata
  • Bumped dependancies

Update your image with:

docker pull promptpanel/promptpanel:1.9

and then re-create your container.

Features

  • Added compatibility with OpenAI's GPT-4o model.
  • Made 4o the default onboarding model for OpenAI.

Enhancements

  • Updated message state to make streaming more fluid on completion.
  • Fix for duplicating threads.

Update your image with:

docker pull promptpanel/promptpanel:1.8

and then re-create your container.

Features

  • Added one-click Gitpod dev environment.

Enhancements

  • Fixes for file upload for plugins (based on new meta field).
  • Tweaks to hello_agent sample template.

Update your image with:

docker pull promptpanel/promptpanel:1.7

and then re-create your container.

Features

  • Added Groq models to the universal LLM chat manifest (lightning fast llama!).

Enhancements

  • Added fixes for incrementally yielding responses.

Features

Open Source:

  • As of v1.6, all of PromptPanel's features are open source.
    • Previous commercial features like user management, access controls, and SSO are now included in OSS.

Agent Plugins:

  • Agent plugins will now default to a chat front-end (no longer needs to be included in plugin directory).
  • manifest.json of plugins now included a capabilities: key which will be fed to templates as context (currently used for enabling / disabling vision on supported models).
  • Deprecated the metadata field across plugins - new field is meta.
    • Use JSON instead of storing as a string and converting (which was initially was used for database compatibility, but we're looking at only supporting Postgres and SQLite for now).
    • Migration script is included on startup to make this change non-breaking (metadata will be removed at a later date).
  • Updated with new plugins across the different default LLM models:

Files:

  • Added file/update endpoint.
  • Added a filename field to data model.
  • New document interface on pre-installed plugins.
    • By default:
      • Agents will prepend document content before the chat history.
      • When context is too long they will truncate the documents passed to the model.
    • We'll be including a RAG use-case as well in an upcoming cookbook item.

Enhancements

  • Using customized textarea field vs Quill dependancy.
  • Fix to allow admin users to use edit_message endpoint.
  • Fixed copy for access controls around application.
  • Fixed toast issue for local models.
  • Tweaks to input sanitation.
  • Minor interfaces tweaks.
  • Reduced AMD64 bloat.

Update your image with:

docker pull promptpanel/promptpanel:1.5

and then re-create your container.

Fixes

  • Fixes issue with 1.4 data-fetching for some plugins

Features

  • Simple starter agents for Cohere and Google Gemini

Enhancements

  • Uses Microsoft's Phi-3 for a better, smaller onboarding model
  • Added Cohere and Gemini to initial onboarding process
  • Updated local quick-models for Ollama to include Command R, Phi, and Qwen

Features

  • OpenID Connect compatibility
  • Added Postgres database compatibility
  • Access controls for Panels (Admin / User / granular Panel logic)
  • Updated auth token handling (prep for API keys), added block list
  • Added server-side logout route
  • Added allowed domain signup and public signup (prep for public sign-up pages)
  • Updated to llama3 in Ollama management and onboarding

Enhancements

  • General visual tweaks and fixes

v1.3 Plugins++

Features

  • Added multi-modal (vision) model agent
  • Updated plugin models to include new OpenAI, Anthropic models (on top of existing universal agent)
  • Added GitPod one-click plugin environment

Enhancements

  • Fixed plugin defaults
  • Updated Ollama error state
  • Fix for truncating hanging sentences
  • General visual tweaks and fixes

Features

  • Added onboarding module (Ollama / API key)
  • Added private: write-only fields in plugin manifest settings

Enhancements

  • General visual tweaks and fixes

Update your image with:

docker pull promptpanel/promptpanel:1.1

and then re-create your container.

Features

  • Added display pictures to Panel creation / edit
  • Updated templates to support mobile displays

Enhancements

  • General visual tweaks and fixes