Find the Best LLM
for Your Project

Compare 100+ models side-by-side with your project's prompts. See real outputs, costs, and quality scores instantly.

No setup required.
No API keys needed.
"We're using models from OpenAI. It's costing us thousands of dollars per month. That new LLM just dropped from Google. Should we switch some of our prompts over?"

How Try That LLM Works

1

Add your prompts

Type them in or import your existing library.

2

Choose your models

Pick specific LLMs or add dozens at a time.

3

Compare results instantly

See differences in tone, reasoning, cost, and quality in one clean grid.

4

Automatically test new LLMs

Every new LLM is automatically tested with your prompts.

1. Add your prompts

  • Type/paste in your prompts or import a CSV
  • Configure your (optional) AI Judges, using your own criteria
  • Use your per-month prompt query counts to track costs
  • No need to re-enter prompts when a new LLM is released
  • See all your prompts in one grid
Fibonacci
15000
prompts/month
Implement a Fibonacci function in TypeScript. Comment the code thoroughly.
LLM Judge Name
LLM doing the judging
OpenAI iconOpenAI: GPT-5 Nano
Judge Criteria. Specify what is a 5 ( = good) and what is a 1 ( = bad ).
Bulk Add every LLM from the past
from
LLMs to add120 new

2. Choose your models

  • Pick specific LLMs or add dozens at a time
  • Sort/filter by family or release date
  • Choose your (optional) AI Judge models
  • No detective work needed to test out the latest models
  • "Let's test every model released in the past 6 months"

3. Compare results instantly

  • See differences in tone, reasoning, cost, and quality in one clean grid
  • Compare model outputs side-by-side
  • Mark your favorite outputs
  • See your AI Judge opinions on the winners
  • Understand which prompts are costing you the most
  • Know whether you're overpaying
  • Understand the cost-quality tradeoffs for each prompt
  • Confidently answer the question "Should we switch models?"
Anthropic: Claude Haiku 4.5
Jan 14, 2026, 10:50:34 PM
$25.68 for 20000 prompts/month
0.0018x o3 Deep Research
or: $1,284 per million prompts
or: 779 prompts per $1
14 input tokens for this prompt
254 output tokens for the LLM's response
$25.68 per month
0.0018x o3 Deep Research

Auto-Add

Email me about Auto-Added LLMs

Favorite LLM Families

4. Automatically test new LLMs

  • Choose to test all, some, or none of any new models with your prompts
  • Your prompts automatically get tested
  • Get an email when the latest LLM responses are ready
  • Never worry about missing a new model
  • Always answer yes to "Have we tried that new model yet?"

Pricing

Choose the plan that fits your project

Free Demo

Honestly it's just a demo, set your expectations low
  • Add your prompts
  • Add your chosen LLMs
  • Auto-add new LLMs
  • Email updates for new LLMs
  • 500 one-time credits
  • Gaze longingly at the Hobbyist plan

Hobbyist

Test your prompts, keep up with the latest LLMs
  • Everything in Free
  • Public or private share links
  • 22,000 credits per month
$24 / month, billed annually
Most Exciting

Pro

You're using it for something real, time to get serious
  • Everything in Hobbyist
  • CSV import/export
  • AI Judges
  • 59,000 credits per month
$49 / month, billed annually

What our users say

See how teams are using Try That LLM to choose the best models for their project.

Frequently asked questions

Wait, can one LLM really cost 1000 times more than another?

Absolutely! Input, Output, and Reasoning token costs can vary wildly between models for the same prompt.

Are we overpaying for our LLM?

Probably! Maybe not for all of your prompts, but likely for some. Compare pricing across models to find the best value.

How do I compare 10+ models at once?

You can add as many LLMs as you want. Want to test your prompts against every model from OpenAI, Anthropic, and Google released in the past 6 months? No problem.

How can I keep up with the latest LLMs?

Daily updates.As new models get released, they'll be automatically tested with your prompts and you'll be notified via email.

Would quality drop if we use a cheaper LLM?

Find out by comparing outputs side-by-side, voting on your favorites, and seeing your winners.

Can I run tests on the results?

Sure! You can run basic comparisons like equals, contains, starts-with, ends-with, etc. for free--they're fast and easy.

Can I use AI to judge the results?

Sure! With a Pro Plan, you can set your own criteria, then choose an LLM to do the judging. You can have multiple judges rating each prompt output, using different judge LLMs and different criteria.

Can you just tell me the best LLM?

Sadly no. But with a Pro Plan, you can hand all your responses off to a Meta Judge LLM. It'll rank all the LLM outputs according to your criteria and you'll get its opinion on the winner.

Can I import my existing prompts?

Sure! With a Pro Plan, you can import your existing prompts from a CSV file. You can choose to import them as drafts or just start prompting with them immediately.

I have a question that's not on this list!

Great! Please get in touch via the Feedback link at the bottom of this page.
Sign UpSee It In Action