Prism is a single workspace for every AI model you use, with your own keys and no middleman taking a cut per message. It supports Anthropic's Claude models, OpenAI's GPT and o-series, Google's Gemini, anything reachable through OpenRouter, and Ollama running locally on your own machine with no key at all. You can add custom model IDs per provider in settings, so a newly released model works the day it ships without waiting for an app update. Chat streams with Markdown and syntax highlighting, extended thinking is shown where the model supports it, and every message carries a token count and a cost estimate so you always know what a conversation is costing. Compare mode is the standout: send one prompt to several models simultaneously and read the answers side by side — the fastest way to decide which model is actually best for a given job. A prompt library keeps reusable system prompts and personas, everything is stored locally as JSON, and API keys are encrypted through the Windows credential store.
What it does
One app for Anthropic Claude, OpenAI GPT and o-series, Google Gemini, OpenRouter and local Ollama
Compare mode sends one prompt to several models at once and shows the answers side by side
Add custom model IDs per provider so new models work the day they launch
Streaming chat with Markdown, syntax highlighting and extended thinking where supported
Per-message token counts and cost estimates so spending stays visible
Prompt library of reusable system prompts and personas
Local-first JSON storage with API keys encrypted through the Windows credential store
Ollama support means you can run entirely offline with no API key at all
Windows installer (.exe) — installs in a couple of clicks, no extra runtimes needed
Platform:
Windows 10/11 (64-bit)
Version:
1.0.0
Download size:
78 MB
Requires:
Bring your own API keys for the providers you want, or run Ollama locally for no key at all. Not code-signed yet, so SmartScreen may warn on first run - choose More info, then Run anyway.
Licence:
No licence key, no trial and no locked features — buy it once and every function is available. You supply your own API keys and pay each provider directly, or run Ollama locally for free.
