Version 1.1.0 – April 21, 2026

What's New

Model providers

OpenRouter
— Use one API key to access many cloud models through OpenRouter’s OpenAI-compatible endpoint.
Local and self-hosted runtimes
— First-class support is now available for
Ollama, LM Studio, vLLM, MLX, and llama.cpp
.
Generic compatible servers
— You can now point Prompt Master Pro at
OpenAI-compatible
or
Anthropic-compatible
HTTP backends.
Expanded model catalog
— Curated model entries were added for major hosted providers including OpenAI, Gemini, DeepSeek, and Claude.
Intercom support
— You can now contact support or report issues directly from the editor sidebar.

Ollama

Added an in-app flow to install or find Ollama, pull allowed models, and use them in Test, Playground, and Deploy.
Support includes official Gemma 4 tags such as
gemma4:e2b
.

Improvements

AI Settings
now has a clearer layout for local providers, OpenRouter, and compatible servers.
Ollama install progress
is less noisy and feels smoother during setup.
Sign-in and account flows
are more consistent across login, registration, profile, and user settings.
Session handling
now restores your last main editor tab more reliably after reload.

Fixes

Playground and Ollama
now display assistant output correctly for models that return reasoning or non-string content.
Evaluations
now correctly resolve model identifiers that contain more than one colon.
Logs
no longer switch you to another tab after resizing the panel.