Inquir Compute · A plan is required to run workloads · Free on request
Long-running AI calls, webhooks and pipelines with scheduled runs, with retries and tracing built in. Start with a function, stay for the application. Node.js, Python, Go or any Dockerfile. No Kubernetes.
$ inquir deploy
✓ Bundling handler 1.2s
✓ Warm runner ready 0.3s
✓ Deployed ai-summarizer (v3)
$ inquir invoke ai-summarizer \
--payload '{"url": "https://techcrunch.com"}'
{
"summary": "AI funding surges as enterprise adoption...",
"wordCount": 824
} SUCCEEDED · 2.4s
THE OLD WAY
Or — just write the function.
Platform Features
One platform for functions, an API gateway, webhooks, pipelines with schedules and container applications. Run full containers and automate recurring work in the same workspace, without managing Kubernetes.
Warm containers keep your functions ready while traffic keeps flowing, so repeat invocations skip the cold start. The very first request after a deploy may still take the cold path — everything after it stays fast.
Every function runs in its own container with no network access until you grant it. Secrets and environment variables are scoped per function, so one workload can never read another's credentials.
Add a route and the function is reachable over HTTPS — with route groups, path parameters and response streaming built in. No ingress to configure, no DNS to wait for.
Add a schedule to a pipeline in the visual editor. Use a cron expression for recurring runs, with retries and execution history in the same workflow.
Write functions in Node.js 22, Python 3.12, or Go 1.22 and bring any npm or pip package. Full container runtimes mean native modules, headless browsers, and ML libraries just work.
Declare openai or @anthropic-ai/sdk in inquir.yaml build.dependencies and add your API key to the function's environment variables. Stream tokens to the browser over server-sent events.
Containers from a Dockerfile: digest-pinned releases, promotion after a health check, several replicas.
How it works
Write a function in the browser, deploy it, add a gateway route and call it over HTTPS. A function needs no Dockerfile.
Write your handler in the browser editor — Node.js, Python, or Go, with syntax highlighting and AI assistance. No local toolchain to set up.
export default Handler(async (ctx) => { const { url } = await ctx.body; return { summary: await ai.summarize(url) }; });
Click Deploy. The platform builds the container and runs your function; add a gateway route and you have an HTTPS endpoint.
$ inquir deploy ✓ Building container 1.2s ✓ Warm container ready 0.3s ✓ Endpoint ready
Call it from anywhere with a standard HTTP request. Any language, plain fetch or curl — no proprietary SDK required.
const res = await fetch(endpoint, { method: "POST", body: JSON.stringify({ url }) });
Follow the code, deployment, and request. Explore each step at your own pace.
Add gateway routes, pipeline schedules, streaming and logs without changing your deployment flow.
package main import ( "context" "encoding/json" "net/http" ) func Handler(ctx context.Context, w http.ResponseWriter, r *http.Request) { var body map[string]any _ = json.NewDecoder(r.Body).Decode(&body) w.Header().Set("Content-Type", "application/json") _ = json.NewEncoder(w).Encode(map[string]any{ "runtime": "go", "echo": body, }) }
Built for LLM
LLM calls are slow, fail sometimes, and are hard to see into. Inquir handles all three — so you ship integrations instead of plumbing.
Minute-long model calls and multi-step chains run to completion — a step's time limit is configurable up to 24 hours.
Every step's input and output, live run status and logs — no observability to wire up yourself.
Background jobs with automatic retries, dead-letter, per-key concurrency and idempotency keys.
Use Cases
From AI agents and webhook processors to background jobs and REST APIs — one platform scales with you, from first deploy to production.
Agent loops, RAG pipelines and token streaming that outlast a normal request: minute-long model calls run to completion, with retries and a trace for every step. The OpenAI or Anthropic SDK is one line in inquir.yaml.
Add a schedule to a pipeline for reports, syncs and cleanup. Cron expressions, retries and run history are part of the pipeline workflow.
Handle Stripe, GitHub, and Slack webhooks reliably: verify signatures, acknowledge fast, and retry idempotently when downstream systems misbehave.
One workspace
A function, a durable job, a pipeline and a container service are four shapes of the same thing: same workspace, same variables, same logs, same team. Reach for the next one when the workload asks for it; nothing gets re-platformed.
Code the platform runs per invocation: HTTP endpoints, webhooks and background jobs with retries. For recurring work, call the function from a pipeline with a schedule.
Run functions on a schedule or connect them into graphs with retries, branching and approvals. Postgres-backed jobs survive restarts, and execution history shows each step.
Any container kept running: raw HTTP and TCP, WebSockets, databases on volumes, private networking. Reach for it when the process must stay up, hold state, or speak a protocol of its own.
Comparison
APIs, webhooks, scheduled pipelines and background jobs — without Kubernetes.
A snapshot based on each provider's public documentation in 2026. Products and limits change often, so check the provider's current docs before you decide. Most of these platforms now offer scheduled runs; the difference is how many separate products you combine. Inquir keeps container apps, functions, pipelines and their schedules in one workspace behind one gateway.
Pricing
Choose Simple or Pro as your project grows, or discuss individual terms with our team. The Free plan is available on request.
The Free plan is enabled on request.
Free access is granted individually, subject to available capacity.
For personal projects and small products ready for their next step.
For teams and production products that need more resources.
For organizations with specific resource, deployment and support requirements.
Free, Simple and Pro differ only in resources. Free is available on request; Enterprise terms are agreed individually.
FAQ
Hi, it's the Inquir Compute team. Ask AI about the product:
Plan resources listed up front. No Kubernetes.
