What do you want to test?
Test my auth flowFind rate-limit friction
Live synthetic run started — 50 agents deployed.
Use it in the browser
Log in, paste a URL or OpenAPI spec, ask a question. Results in minutes.
Synthetic users for API testing
Get rid of user friction in your app.
Register your product, configure test journeys, and get ranked friction clusters — from the same engine that powers multi-agent population simulation.
Beta testers from YC and A16Z
Most API changes ship to users you've never tested with.
Discovery
Research
Interviews
Synthesis
Research takes weeks you don't have.
You ship
Week 1
too late...
User data
Week 3+
Feedback arrives after you've shipped.
Assumptions get baked in, never tested.
Will auth fail silently?
Is rate limiting clear?
Do users understand errors?
Is the flow obvious?
Will they find the docs?
Are we solving the right thing?
How it worksFeedback in minutes.
Feedback in minutes.
Here's how it works.
What do you want to test?
Ask a question
Type what you want to understand. Attach an OpenAPI spec or staging URL for context.
Deploying agents
AM
SL
JK
Activate user movement
AI personas navigate your API independently and capture every interaction.
Issues found
UX-01Auth timeout unclear for new usersTO DO
UX-02Rate limit response missing Retry-AfterIN PROGRESS
UX-03Validation error format inconsistentTO DO
Get clear answers
See exactly where users struggled and what to fix before you ship.
One tool, three surfaces
Available where you work
$ pip install omnisim
$ omnisim run --product prod_abc
50 personas deployed
12 interactions captured
3 friction issues found
Run it from your terminal
Two commands to your first test. Pipe results into any workflow or CI pipeline.
Cursor→OMNISIM
omnisim.run_test(url)
Deploying 50 personas…
12 interactions captured
omnisim.get_findings()
3 issues found
Wire it as an MCP server
Add OMNISIM to Cursor or Claude Code. Test without leaving your editor.
Intelligent features
Make decisions with confidence
Friction clusters
Issues are automatically grouped by severity and component. See exactly where users struggled and what to fix before you ship.
API-01Auth timeout unclear
API-02Rate limit response missing headers
API-03Validation error format inconsistent
Pricing
Three tiers. One credit pool.
Free
$0/mo
- $20 credits / mo (1 credit = $1)
- Up to 5 agents · 1 round per run
- ~20 behavioral runs / mo at max agents
- 5 API calls / day
- 1 registered product
Most popular
Pro
$30/mo
Launch-grade synthetic testing — 40 AI users per run on your staging API or UI.
- 40 AI agents per run · up to 3 rounds
- $30 credits / mo (~15 full runs at max agents)
- Advanced agents + enhanced DOM screen reads
- 5 products · webhooks · 25 API calls / day
Pro+
$60/mo
Scale synthetic coverage — 80 agents, 4 rounds, and room for multiple products.
- 80 AI agents per run · up to 4 rounds
- $60 credits / mo (~15 full runs at max agents)
- Advanced agents + enhanced DOM screen reads
- 20 products · webhooks · 100 API calls / day
1 credit = $1 · Enterprise
Waitlist
Join for early access
We'll reach out when your account is approved.
