LLM workflows, without orchestration code.
Design workflows visually. Run them via a simple API.
Works with:
See example Start building
Prospecting Workflow Demo
Hover nodes to see prompts & outputs
Restart workflow
Hover nodes to see their content
Input
Filter (Marketing)
1.2s
gpt-4o-mini
A/B test models, prompts, and parameters
Enrich A (Fast)
1.8s
gpt-4o
Enrich B (Detailed)
1.8s
claude-3-5-sonnet
Plan Outreach
gpt-4o
Output
Press enter or space to select a node. You can then use the arrow keys to move the node around. Press delete to remove it and escape to cancel.
Press enter or space to select an edge. You can then press delete to remove it or escape to cancel.
Scroll away to exit
Why RunStack?
LLM orchestration doesn't scale in code.
Building production LLM systems in code becomes brittle fast.
Without RunStack
Orchestration in code
Every prompt change requires redeploy
Testing variants means branching production code
Choosing the best model requires code changes.
Result: Fragile & slow
With RunStack
Visual orchestration
Design workflows visually, no code changes
Update prompts and models instantly
Test variants side-by-side before shipping
Execute via simple API with only 2 endpoints
Minutes instead of days to ship
Result: Fast & reliable
One requires constant maintenance. The other ships faster.
Features
Everything you need to run LLM workflows
From visual design to execution, built for iteration.
Visual workflow builder
Design, edit, and run LLM workflows without touching production code.
Input
Stage 1
Variant 1
gpt-4o
Variant 2
claude-3-5-sonnet
Variant 3
mistral-large
Stage 2
Variant 1
gpt-4o-mini
Stage 3
Variant 1
deepseek-r1
Variant 2
gemini-2.5-flash
Output
Press enter or space to select a node. You can then use the arrow keys to move the node around. Press delete to remove it and escape to cancel.
Press enter or space to select an edge. You can then press delete to remove it or escape to cancel.
Safe iteration
Deploy version in 2 clicks.
Your browser does not support the video tag.
Variants by default
Test prompts and models side-by-side.
Variant 1
Take every entries from ${{stage1}} and provide basic company info: industry and 2 talking points.
gemini-2.5-flash
Variant 2
For each prospect in ${{stage1}}, provide detailed company info: industry, size, and 3 specific talking points with data.
claude-3.5-sonnet
Simple API
Execute workflows with only two endpoints.
POST /v1/workflows/:id/run → execute workflow
GET /v1/runs/:runId → fetch status and output
Pricing
Simple, Transparent Pricing
Choose the plan that works for you All plans include access to our platform.
Free
€0123456789/ month
Forever free
- Build and edit workflows
- Publish workflows
- API access (limited)
- 50 execution credits / month
- 1 concurrent run
- 1 API key
- Rate limit: 1 req/sec
Start Free
Get started with basic features
Popular
Starter
€01234567890123456789/ month
Billed monthly
- Everything in Free
- 1000 execution credits / month
- 2 concurrent runs
- 3 API keys
- Rate limit: 3 req/sec
Upgrade to Starter
For individuals and small projects
Pro
€01234567890123456789/ month
Billed monthly
- Everything in Starter
- 3000 execution credits / month
- 5 concurrent runs
- 10 API keys
- Rate limit: 10 req/sec
Upgrade to Pro
For professionals with custom models
Scale
€012345678901234567890123456789/ month
Billed monthly
- Everything in Pro
- 8000 execution credits / month
- 10 concurrent runs
- 25 API keys
- Rate limit: 30 req/sec
Upgrade to Scale
For teams with high volume needs
FAQ
FAQ
Common questions about RunStack