DURABLE EXECUTION · MULTI-MODEL · RAG-READY

AI agents that survive production.

The visual canvas for durable AI workflows. Drag, connect, deploy. Durable execution handles the retries, the long-running jobs, and the cost tracking you'd otherwise build yourself.

Runs in your browser. Nothing to install.
Built on Trigger.dev's durable execution Multi-model: Claude, GPT, Gemini, and more Cost tracked on every run No infrastructure to run
THE PROBLEM

Most AI agents work once.

They work in the demo. They work in the screenshot. They work until an API times out, or the model returns an empty string, or the job runs for more than 30 seconds, or someone tries to actually use the thing at scale. Then they stop working.

Making an AI agent that holds up in production is a second, much harder job — vector databases, retry logic, async queues, cost tracking, observability. The people who understand what AI can do for their business shouldn't have to build all of that themselves.

PRIMITIVE 01 / FABRIC

A visual canvas for AI agents that hold up.

You describe the workflow. Knitch runs it. Every step is checkpointed, every failure retries automatically, every token is tracked. The agents you build on the canvas run for hours and cost what you expect.

weaves/ doc-extract-and-route / Agent · Extract
live · last run 12s ago
Mailhook
support@acme.co
Vector KB
policies · 42 docs
Agent · Extract
claude-sonnet-4.6
toolslookup, classify
cost$0.0081
Linear · Create
team: support
Draft Reply
gmail · sandbox
WHAT'S IN THE BOX

Five things that, together, nothing else has.

Pick any one and someone sells it. Stitched together, they're a workflow you can actually ship.

Agent in minutes, not weeks

A working agent takes under ten minutes. Webhook input, AI node with agent mode, output. No infrastructure to configure. The hard parts are already built.

// under 10 minutes Webhook Agent Output ✓ live

RAG, ready on day one

Knowledge bases, vector search, hybrid retrieval, all pre-wired. Drop in documents or connect a source and the agent can search and reason over your content. No pgvector. No embedding pipeline.

Upload "policies.pdf" → 1,284 chunks · hybrid index

Multi-model by design

Not an OpenAI wrapper. Assign Claude, GPT, Haiku, Gemini, or an open model per node. The right model for each step, with the cost and latency you picked for a reason.

classify haiku · draft sonnet-4.6 · score gpt-4o-mini

Multi-agent orchestration, built in

A Dynamic Orchestrator that takes a goal and spawns sub-agents to handle it. Parallel fan-out, synthesis, evaluator loops. Not on the roadmap — shipping today.

orchestrate("draft quarterly report") → 4 sub-agents · fan-out · synth

Every token accounted for

Every run produces a full trace. Cost by node, tokens in and out, latency per step, tool calls made, embeddings retrieved. Point at the specific LLM call that's expensive and decide what to do about it.

trace #4,281 · $0.0042 · 1,284 tok · 2.3s → 0.8s → 1.1s
THREE WAYS TO BUILD

From one agent to many.

Every weave on Knitch fits into one of three shapes. Start with the simplest that still solves the problem.

agent

Single Agent

One AI node. It reasons, picks tools, loops until done. Best for focused tasks: triage, classification, extraction, generation.

Multi-Agent Workflow

Multiple AI nodes in sequence or branching. Each handles a phase and passes structured output to the next. Best for pipelines with distinct stages.

orch

Dynamic Orchestration

One Orchestrator node takes a goal, spawns sub-agents to handle it, and synthesizes the result. Best for complex tasks where sub-tasks can't be predicted up-front.

THE CANONICAL DEMO

A service desk that triages its own email.

An email arrives. An agent reads it, classifies the issue, routes it to the right team, creates the ticket, drafts a response. Autonomously. With full cost tracking. Using a different model for each step.

Built on Knitch in eight nodes and about seven minutes of wiring. The full weave is below.

weaves/ service-desk-triage / run #4,293
replaying · 2.8s elapsed · $0.0063
Mailhook
support@acme.co
Customer KB
12,430 docs
Classify
claude-haiku-4.5
cost$0.0003
Route · Severity
P0 / P1 / P2
Draft Reply
sonnet-4.6
cost$0.0054
Linear · Ticket
team: support
Gmail · Reply
awaiting approval
Return
200 · OK
WORKFLOW COPILOT

A copilot that already read your whole workflow. Ask it anything.

Open it on the canvas and it already knows your whole workflow — every node, your knowledge bases, your stored data, and what each node's last run cost. Ask why a step is slow, which model a node should use, or where a number came from. It searches with the same retrieval your agents use, points to the exact node it means, and tells you what to change and why. It reads and explains, it doesn't touch your workflow.

weaves/ service-desk-triage / canvas
copilot · ⌘I
Mailhook
support@acme.co
Classify
claude-haiku-4.5
cost$0.0003
Customer KB
12,430 docs
Draft Reply
claude-sonnet-4.6
cost$0.0054
Workflow Copilot · model choice
@ Classify ×
Classify just tags each email's topic and urgency. What model should it use?
inspect_node · Classify
@Classify only does short, structured labeling, so it doesn't need sonnet-4.6. haiku-4.5 is a better fit, near-identical accuracy on classification at a fraction of the cost, roughly $0.0003 versus $0.0036 per call. On a high-volume node like this, that difference is most of its spend. Open Classify's panel and switch its model to haiku-4.5, and keep sonnet on @Draft Reply, where wording quality matters.
8 nodes 12,430 KB docs 3 storage 5 tools
Ask about this workflow… 4,210 tokens · $0.0089
Same brain as your agents

It answers using the exact models, retrieval, and tools your nodes run on — so what it tells you matches what the workflow actually does.

It points, you click

Every node it names is a live chip. Click @Draft Reply and the canvas centers on that node.

It saw the last run

Tokens, cost, and what each step did — pulled from your real executions, not guessed. Ask where a number came from and it shows you.

WHERE KNITCH FITS

Workflow tools, developer tools, and the middle nobody else is serving.

Zapier can't do agents. Trigger.dev wants you to write TypeScript. Knitch is for the person in between.

Zapier · Make Trigger.dev · Inngest Knitch
Who builds it Anyone Developers only AI-aware, not a developer
AI capability Basic AI steps Full control, code required Full agentic capability, no code
RAG / knowledge Not available Build it yourself Pre-built, day one
Multi-agent Not available Build it yourself Built in, visual
Model choice Limited Full control Multi-provider, per node
Cost observability Minimal Depends Full trace per run
Time to first agent N/A Days to weeks Under ten minutes
OBSERVABILITY

See what every agent did. And what every agent cost.

Every run produces a complete trace: cost breakdown by node, tokens per agent and sub-agent, latency per step, tool calls made, embeddings retrieved. No black box. When the bill shows up, you'll know why.

Mailhook 0.0s Customer KB 0.2s · 12 chunks Agent · Classify 0.8s · $0.0003 Linear 1.1s Gmail 0.3s

Run #4,293 · metrics

Input tokens2,418
Output tokens1,182
Total tokens3,600
Cost$0.0063
Latency (p50)2.8s
Retries0
Tools invokedkb.lookup, linear.create

Stitches

mailhook.in0.0s
kb.lookup — 12 chunks0.2s
agent.classify — haiku-4.50.8s
route.severity → P10.01s
agent.draft — sonnet-4.61.1s
linear.create — ticket ACME-48210.6s
gmail.draft — awaiting approval0.3s
UNDER THE HOOD

The boring parts, already handled.

You focus on what the agents should do. Knitch handles the rest.

Durable execution

Every node checkpointed. Failures retry automatically with exponential backoff. Long-running jobs survive deploys and crashes — a weave that starts on Monday can finish on Thursday.

Six agent patterns, ready to drop in

Classify & Route. Extract & Transform. Agentic Loop. Knowledge-Augmented. Parallel Orchestration. Persistent Chat Agent. Start from a pattern or build from scratch.

Webhook, schedule, mailhook, chat

Four trigger types. No cron daemon to run. No server to SSH into. Point a URL at Knitch and an agent picks up from there.

Bring your own keys

Use your Anthropic, OpenAI, Google, xAI, DeepSeek, or Mistral account. Knitch passes through. You pay what the model costs — nothing stacked on top per token.

PRICING

Start free. Pay for execution, not seats.

Usage-based, tracked by the same telemetry that powers cost observability. The bill matches the trace.

Free

Build, test, and ship your first agent. Generous limits on runs, tokens, and knowledge base size. Everything you need for a real proof of concept.

Start free →

Enterprise

SSO and SAML, audit logs, RBAC, bring-your-own-key, dedicated support, and a custom SLA. Talk to us.

Talk to us →

See the full tier breakdown on the pricing page.

Build one. See for yourself.

Ten minutes to a working agent. No credit card. No infrastructure.

Start free → Read the quickstart
No install. Works in any modern browser.