Many heads LLMs are better than one

Chat with many LLMs at once, and combine responses.

Tweak prompts and other settings for the perfect output.

Try it for free See it in action

Find your perfect output

Four ways to get better answers from AI.

Compare models. Pit Claude against GPT against Gemini in seconds—you'll be surprised how different they are.

Compare prompts. Tweak your wording until you find the magic phrase that unlocks dramatically better answers.

Research mode. Need fresh info? AI searches the web and brings back citations, not stale training data.

Custom mode. Ready to geek out? Adjust temperature, reasoning depth, and more.

45+ LLMs supported

And more coming!

Claude 3.5 Sonnet

Claude 3.7 Sonnet

Claude 3.7 Sonnet (Thinking)

Claude 3.5 Haiku

Claude 4 Sonnet

Claude 4 Sonnet (Thinking)

Claude 4 Opus

Claude 4 Opus (Thinking)

Claude 4.5 Sonnet

Claude 4.5 Sonnet (Thinking)

Claude 4.5 Haiku

Claude 4.5 Haiku (Thinking)

GPT-4o

GPT-4o mini

GPT-4.1

GPT-4.1 mini

o3 (Thinking)

o3 mini (Thinking)

o4 mini (Thinking)

GPT-5

GPT-5 Mini

GPT-5 Nano

GPT-5.1

Gemini 3 Pro (Preview)

Gemini 2.5 Pro (Thinking)

Gemini 2.5 Flash

Gemini 2.5 Flash (Thinking)

Gemini 2.5 Flash Lite

Gemini 2.5 Flash Lite (Thinking)

Gemini 2.0 Flash

Grok 4 Fast (Thinking)

Grok 4 (Thinking)

Grok 3

Grok 3 Mini (Thinking)

Grok 2 Vision

DeepSeek V3.2 Speciale

DeepSeek V3.2

DeepSeek V3.1 (Thinking)

DeepSeek V3 (Chat)

DeepSeek R1 (Thinking)

Mistral Medium 3.1

Mistral Large 2.1

Magistral Medium 1.2 (Thinking)

Pixtral Large

Mistral Saba

Perplexity Sonar

Perplexity Sonar Pro

Perplexity Sonar Reasoning Pro

Kimi K2

Kimi K2 (Thinking)

Qwen 3 Max

Qwen 3 VL 235B (Thinking)

Qwen 3 235B Instruct

Qwen 3 Coder 480B A35B Instruct

Qwen 2.5-VL 32B Instruct

GLM 4.5

GLM 4.6

Cogito v2 Llama 405B

Cogito v2 DeepSeek 671B MoE

GPT OpenSource 20B

GPT OpenSource 120B

MiniMax M2

Share, learn, improve

AI gets better when you don't go it alone.

Share your work. Found a killer prompt? Discovered which model aces your use case? One click generates a shareable link for teammates, clients, or your future self. Privacy-first—revoke access anytime.

Clone and remix. Every shared evaluation can be duplicated and tweaked. Learn from how others phrase their requests, then make it your own.

Prompt Doctor. Stuck? Our AI assistant analyzes what's not working and suggests improvements—like having an expert prompt engineer on call, 24/7.

Get started now See it in action

Watch the demo

A full walkthrough of PromptHQ and how you can use it.

Frequently asked questions

Have more questions or just wanna talk prompts and evals? Email us at help@prompthq.run

Is my information private?

Yes. We only use the information you submit to serve and improve the product. We will not sell or distribute your data. You can share your data with people you choose through our collaboration features.

Do you train your LLM on my data?

No. Your data belongs entirely to you.

What is your privacy policy?

You can find our privacy policy here: https://www.prompthq.run/legal/privacy

Why should I use this app over apps like ChatGPT or Gemini?

PromptHQ is not a replacement for ChatGPT or Gemini, or even an AI chat app. It is a tool to help you use LLMs better — by evaluating and iterating on different LLMs and prompts and settings like temperature, thinking budgets, etc.

Why is AI fluency important?

The AI revolution is like the computer revolution — if you stuck with pen and paper and faxes, you got left behind. Read more about how the leading companies are already asking their employees to get AI-fluent.

Is prompting LLMs really that hard?

It is different from Google Search where you can write a terse query but do most of the work afterwards trying to make sense of the search results. With LLMs, a little upfront work in writing a good prompt can make it do all the work for you.

Who is the team behind PromptHQ.run?

PromptHQ.run is operated by the team at GnoDesk. This was one of our most useful internal tools and we decided to make it publicly available.