ProteinIQ API

Research with the ProteinIQ bioinformatics API

Run 100+ bioinformatics tools from code. One endpoint, async jobs, and signed result files, ready for your pipeline.

curl -X POST https://proteiniq.io/api/v1/jobs \
  -H "Authorization: Bearer $PROTEINIQ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "tool": "esmfold",
    "name": "Lysozyme ESMFold run",
    "input": {
      "inputs": [{
        "id": "seqs_1",
        "slotId": "protein",
        "kind": "protein",
        "format": "fasta",
        "content": ">lysozyme\nKVFGRCELAAAMKRHGLDNYRGYSLG",
        "source": { "type": "text" }
      }]
    },
    "settings": {}
  }'

243+

Tools, one contract

Every tool runs through the same submit, poll, and result loop.

B200

GPU-backed compute

Jobs run on NVIDIA B200, H200, and A100 GPUs. Minutes, not days.

REST

No SDK required

Plain HTTP and JSON with a bearer token. Use any language.

Built for production pipelines

A small, predictable API surface for notebooks, services, and lab systems.

One API, every tool

Inspect a tool, then run it with the same request shape. Folding, docking, and validation all share one job model.

Async jobs and live events

Submission returns instantly. Poll for status or stream server-sent events until the job finishes.

Structured results and files

Results return structured output plus a manifest of signed download URLs for PDBs, CSVs, and reports.

Quote before you run

Estimate credits with a quote, then cap spend per job with a reserved-credit budget. No surprises.

OpenAPI and typed clients

A machine-readable OpenAPI document describes every route, so you can generate typed clients.

Safe retries

Send an idempotency key and retries reuse the same job instead of creating duplicates.

From request to result

Submit a job, poll until it finishes, download the files. The same steps in any language.

Full quickstart

# 1. Submit a job
JOB_ID=$(curl -s -X POST \
  -H "Authorization: Bearer $PROTEINIQ_API_KEY" \
  -H "Content-Type: application/json" \
  https://proteiniq.io/api/v1/jobs \
  -d '{"tool":"esmfold","name":"Fold","input":{"inputs":[{"id":"seqs_1","slotId":"protein","kind":"protein","format":"fasta","content":">s\nMKT...","source":{"type":"text"}}]},"settings":{}}' \
  | jq -r '.id')

# 2. Poll until the job reaches a terminal status
curl -s -H "Authorization: Bearer $PROTEINIQ_API_KEY" \
  https://proteiniq.io/api/v1/jobs/$JOB_ID/status

# 3. Fetch the result and download the first file
curl -s -H "Authorization: Bearer $PROTEINIQ_API_KEY" \
  https://proteiniq.io/api/v1/results/$JOB_ID -o result.json
curl -L "$(jq -r '.files[0].url' result.json)" -o result-file

Connect through MCP

Point Codex, Claude Desktop, Cursor, or another MCP client at the remote endpoint. The same API key authorizes agent calls.

Agents can inspect tools, submit jobs with submit_job, poll status, fetch results, and read JSON resources with resources/read.

MCP config

{
  "mcpServers": {
    "proteiniq": {
      "type": "http",
      "url": "https://proteiniq.io/mcp",
      "headers": {
        "Authorization": "Bearer ${PROTEINIQ_API_KEY}"
      }
    }
  }
}

A small, stable surface

Six route groups cover the full lifecycle, from inspecting a tool to downloading results.

Account

GET /api/v1/account

Read workspace credits, concurrency, rate-limit policy, and key metadata.

Tools

GET /api/v1/tools

List available tools and inspect input contracts, settings, and credit models.

Jobs

POST /api/v1/jobs

Submit work, list jobs, poll status, or cancel queued runs.

Events

GET /api/v1/jobs/{id}/events

Stream server-sent job status updates instead of polling.

Results

GET /api/v1/results/{id}

Fetch structured output and signed file download URLs.

OpenAPI

GET /api/v1/openapi.json

Pull the machine-readable contract to generate typed clients.

Bring ProteinIQ into your pipeline

Create a key from a paid workspace and run your first job in minutes. No installs, no GPU setup.

Get an API key Contact sales

curl -X POST https://proteiniq.io/api/v1/jobs \ -H "Authorization: Bearer $PROTEINIQ_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "tool": "esmfold", "name": "Lysozyme ESMFold run", "input": { "inputs": [{ "id": "seqs_1", "slotId": "protein", "kind": "protein", "format": "fasta", "content": ">lysozyme\nKVFGRCELAAAMKRHGLDNYRGYSLG", "source": { "type": "text" } }] }, "settings": {} }'

# 1. Submit a job JOB_ID=$(curl -s -X POST \ -H "Authorization: Bearer $PROTEINIQ_API_KEY" \ -H "Content-Type: application/json" \ https://proteiniq.io/api/v1/jobs \ -d '{"tool":"esmfold","name":"Fold","input":{"inputs":[{"id":"seqs_1","slotId":"protein","kind":"protein","format":"fasta","content":">s\nMKT...","source":{"type":"text"}}]},"settings":{}}' \ | jq -r '.id') # 2. Poll until the job reaches a terminal status curl -s -H "Authorization: Bearer $PROTEINIQ_API_KEY" \ https://proteiniq.io/api/v1/jobs/$JOB_ID/status # 3. Fetch the result and download the first file curl -s -H "Authorization: Bearer $PROTEINIQ_API_KEY" \ https://proteiniq.io/api/v1/results/$JOB_ID -o result.json curl -L "$(jq -r '.files[0].url' result.json)" -o result-file