CrowdSorcerer is an AI task API that handles web research, entity enrichment, document parsing, code execution, PII detection, and more -- all through a single unified endpoint.
Submit a task. Get structured results back. No infrastructure to manage.
TypeScript:
import { CrowdSorcerer } from "@crowdsourcerer/sdk";
const client = new CrowdSorcerer({
apiKey: process.env.CROWDSOURCERER_API_KEY!,
});
// Submit a task and wait for the result
const task = await client.webResearch(
{ query: "latest AI developments 2026", max_results: 5 },
);
console.log(task.status); // "completed"
console.log(task.output); // structured research resultsPython:
import os
from crowdsourcerer import CrowdSorcerer
client = CrowdSorcerer(api_key=os.environ["CROWDSOURCERER_API_KEY"])
# Submit a task and wait for the result
task = client.tasks.create(
type="web_research",
input={"query": "latest AI developments 2026", "max_results": 5},
)
result = client.tasks.wait(task.task_id)
print(result.status) # "completed"
print(result.output) # structured research resultscurl:
curl -X POST https://crowdsourcerer.rebaselabs.online/v1/tasks \
-H "Authorization: Bearer $CROWDSOURCERER_API_KEY" \
-H "Content-Type: application/json" \
-d '{"type": "web_research", "input": {"query": "latest AI developments 2026", "max_results": 5}}'- Create an account at crowdsourcerer.rebaselabs.online/register -- every account gets 1,000 free credits to start.
- Generate an API key from your dashboard.
- Set the environment variable:
export CROWDSOURCERER_API_KEY="cs_live_your_key_here"
- Install the SDK (or use curl directly):
# TypeScript / Node.js npm install @crowdsourcerer/sdk # Python pip install crowdsourcerer-sdk
- Run an example from this repo.
| Task Type | Description | Example Use Case |
|---|---|---|
web_research |
Search the web and return structured results | Market research, competitive analysis |
entity_lookup |
Enrich entity data (companies, people) | Lead enrichment, CRM data quality |
document_parse |
Extract text, tables, and structure from documents | Invoice processing, contract analysis |
llm_generate |
Generate text using large language models | Content creation, summarization |
pii_detect |
Find and mask personally identifiable information | Compliance, data anonymization |
screenshot |
Capture web page screenshots | Monitoring, archival, testing |
code_execute |
Run code in a sandboxed environment | Data processing, calculations |
data_transform |
Convert between data formats (CSV, JSON, etc.) | ETL pipelines, data migration |
audio_transcribe |
Transcribe audio files to text | Meeting notes, podcast indexing |
web_intel |
Extract metadata, links, and content from URLs | SEO analysis, link auditing |
| File | Description |
|---|---|
basic-task.ts |
Create a single task, poll for results |
web-research.ts |
Web research with webhook delivery |
batch-processing.ts |
Submit multiple tasks in batch |
| File | Description |
|---|---|
basic_task.py |
Create a single task, poll for results |
web_research.py |
Web research with webhook delivery |
batch_processing.py |
Submit multiple tasks in batch |
| File | Description |
|---|---|
examples.sh |
All core API operations as curl one-liners |
Base URL: https://crowdsourcerer.rebaselabs.online
| Method | Endpoint | Description |
|---|---|---|
POST |
/v1/tasks |
Create a task |
GET |
/v1/tasks/{id} |
Get task status and result |
POST |
/v1/tasks/batch |
Create up to 50 tasks at once |
GET |
/v1/tasks |
List your tasks (with filters) |
DELETE |
/v1/tasks/{id} |
Cancel a pending task |
GET |
/v1/credits/balance |
Check your credit balance |
Full API documentation: crowdsourcerer.rebaselabs.online/docs
All requests require a Bearer token in the Authorization header:
Authorization: Bearer cs_live_your_key_here
API keys are managed from your dashboard. You can create multiple keys with different scopes for different environments.
CrowdSorcerer can deliver task results to your server via webhooks. Pass a webhook_url when creating a task:
{
"type": "web_research",
"input": {"query": "AI trends 2026"},
"webhook_url": "https://your-server.com/webhook"
}Webhook payloads are signed with HMAC-SHA256 for verification. See the web research examples for implementation details.
- Website: crowdsourcerer.rebaselabs.online
- API Docs: crowdsourcerer.rebaselabs.online/docs
- Sign Up: crowdsourcerer.rebaselabs.online/register
- TypeScript SDK: @crowdsourcerer/sdk on npm
- Python SDK: crowdsourcerer-sdk on PyPI
MIT -- see LICENSE.