Your prompt is the ceiling on every answer you get.
Researchers have published thousands of prompting techniques that measurably improve AI output. You know maybe five of them. GetDecision searches all 5,819, has ten AI specialists argue over the best fit for your exact task, and hands you a ready-to-run prompt built on the winner.
Free to join · your first search is on us.
This is what a decision looks like
Real screens from the live app. No mockups.
The verdict view: three techniques ranked by fit score, with strengths, considerations and when to use each. You pick, it never picks for you.
Plain language in, any of 97 languages. Files welcome.
Every search, choice and generated prompt is saved, searchable and organisable into collections.
The advice "just write better prompts" is useless.
You've seen it happen. The same model that writes brilliant analysis for a colleague gives you a wall of generic filler.
The difference is not the model. It's the technique. Chain-of-thought, tree-of-thoughts, persona scaffolds, decomposition patterns, self-consistency checks. Hundreds of papers have measured what works for which kind of task.
Nobody has time to read them. So most people recycle the same three habits and accept whatever comes back.
What happens in the ninety seconds
after you hit search
You describe the task.
Plain language, any of 97 languages. Attach files if you want: PDF, Word, Excel, images, up to 10 of them. The system figures out what kind of thinking your task needs.
It searches the way a research librarian would.
Your intent gets translated into the vocabulary of the research literature, then matched against 5,819 techniques indexed from published papers. Filters relax themselves if they're too strict. Near-duplicate techniques get pushed apart, so the shortlist holds ten different ideas instead of the same idea ten times.
Ten specialists read the actual research.
Each of the top ten candidates gets its own dedicated AI agent. One agent, one technique, the full source document behind it, about 5,000 words each. No skimming, no summary-of-a-summary.
A judge picks the top three and shows its work.
A reasoning model with a 15,000-token thinking budget compares all ten reports and ranks the top three: strengths, weaknesses, when to use each, when to avoid. You see why. You pick. It never picks for you.
You get a prompt built for your task.
A complete prompt that already applies the chosen technique to your specific task, ready to paste into whichever of 7 models you choose, from budget to premium. No blanks to fill in.
Built like a review panel,
not a search box
It reads the whole paper.
Most tools match your query against titles and tags. GetDecision loads the full research document behind each candidate technique before scoring it, about 5,000 words per candidate.
One agent per document. On purpose.
Long contexts make models lose details in the middle. So every candidate gets a dedicated specialist with nothing else to think about. Ten agents run in parallel.
We turned off a feature because it was flaky.
Multi-query expansion looked great on paper. Our tests showed it made results inconsistent between runs, 20 to 60 percent variance, with no recall gain. It's off. When we test our own pipeline, you get the results.
You see every decision the pipeline makes.
Which of the 5,819 candidates survived each stage. Why the losers got dropped. What each step cost, in real tokens and real dollars. The whole funnel is inspectable, live.
Watch your search think
Every search produces a full trace: seven pipeline stages, per-stage inputs and outputs, token counts, and the exact cost, usually around 10 cents. A live diagram shows candidates flowing from 5,819 down to your top 3.
If a recommendation looks off, you don't have to trust it or drop it. You can open the trace and see where the reasoning went.
Everything around the pipeline works today
Session history
Every search, choice, and generated prompt saved and searchable.
Favorites and collections
Organize techniques by project. Drag and drop.
File-aware prompting
Your uploaded documents shape both the search and the final prompt.
Model choice
7 models across budget, standard, and premium tiers. Prices synced from OpenRouter.
Regenerate with another technique
Got your prompt? Rebuild it with any other technique from your Top-3 and compare the results side by side.
On the bench right now
Consilium: several models answer independently, then debate to consensus.
Deeper context enrichment for ambiguous tasks.
● We only list what's built. Everything above this section works today.
Simple monthly plans
- One full run: search + generated prompt
- All 7 generation models
- Full history & collections
- No card required
- ≈ 30 full runs on premium models
- ≈ 40 runs with Claude Sonnet 5
- Credits never expire
- Same features as every plan
- ≈ 90 full runs on premium models
- ≈ 120 runs with Claude Sonnet 5
- Credits never expire
- Same features as every plan
- ≈ 240 full runs on premium models
- ≈ 320 runs with Claude Sonnet 5
- Credits never expire
- Same features as every plan
A full run = one search (20 credits: 5,819 techniques → ten specialists → judge → your Top-3) plus one prompt generation on a premium model (12–13 credits). Cheaper models generate for 1–5 credits. Your plan refreshes your credits monthly, and you can cancel anytime.
Fair questions
Is this a prompt library?
No. Libraries give you 10,000 prompts and leave the choosing to you, which is the hard part. GetDecision does the choosing, with evidence, and then writes the prompt.
Why does a search take about a minute and a half?
Because ten agents are reading full research documents and a judge is comparing their reports. It's the difference between asking a friend and convening a panel. For a decision that matters, ninety seconds is cheap. And you watch every stage live while it runs.
Which models does it use?
Gemini 2.5 Flash-Lite runs scoring and the ten specialists, Claude Sonnet 4.5 with extended thinking is the judge. For the final prompt you pick from 7 models: GPT-5.6 Sol, Claude Opus 4.8, Claude Sonnet 5, Gemini 3.1 Pro, Grok 4.5, and two budget options.
What does it cost?
Your first search is free. After that, monthly plans from $10 refresh your credit balance to the plan allowance every month. A search costs 20 credits, generating the final prompt adds 1–13 depending on the model you choose. Subscribers can buy one-time top-up packs inside the app. Cancel anytime.
Where do the 5,819 techniques come from?
From published AI research. Each technique keeps its source reference, and the full extracted document travels with it through the pipeline.
The next hard question you ask deserves better than a guess.
Describe your task. Watch 5,819 techniques become 3. Take the prompt.
Launch App