# Parallel Task MCP

Connects your agent to hosted deep web research and batch dataset enrichment. Ideal for sourcing facts and company profiles without leaving the editor.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 16
- Author: parallel-web
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/parallel-web/task-mcp
- Updated: 2026-09-11
- HTML: https://affpapa.org/en/nevibecode/mcp/parallel-task-mcp

## What it does

The server exposes Parallel's cloud-hosted research infrastructure to MCP-compatible editors. It supplies four tools: `createDeepResearch` for sourced investigations, `createTaskGroup` to fill out structured schemas across lists of items, plus `getStatus` and `getResultMarkdown` to monitor runs and fetch Markdown summaries.

Work runs asynchronously in the cloud. You do not need to host anything on your machine. Your assistant starts a job, retains the run ID, and polls for completion while you continue coding.

Tasks are billable through your Parallel account, so missing credits or invalid authentication will trigger an HTTP 402 status.

## Requirements
- Remote server: `https://task-mcp.parallel.ai/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http task-mcp https://task-mcp.parallel.ai/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add parallel-task-mcp --url https://task-mcp.parallel.ai/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "parallel-task-mcp": {
            "url": "https://task-mcp.parallel.ai/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "parallel-task-mcp": {
            "type": "http",
            "url": "https://task-mcp.parallel.ai/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http parallel-task-mcp https://task-mcp.parallel.ai/mcp
```

## When to use it
- produce detailed topic research with verified source citations
- enrich spreadsheets or item batches using a strict schema
- run long web research asynchronously in the background
- pull formatted research summaries directly as Markdown

## Example prompts
- Run `createDeepResearch` on Parallel Web Systems to summarize its product offerings with sources.
- Use `createTaskGroup` to look up the founding year and headquarters for a list of three tech companies.
- Check task progress using `getStatus` and fetch the final report with `getResultMarkdown`.

## Similar
- [parallel-web](https://affpapa.org/en/nevibecode/skills/parallel-web.md) — skill · 2000 installs
- [parallel-deep-research](https://affpapa.org/en/nevibecode/skills/parallel-deep-research.md) — skill · 15600 installs
- [migrate-to-parallel](https://affpapa.org/en/nevibecode/skills/migrate-to-parallel.md) — skill · 3200 installs
- [dispatching-parallel-agents](https://affpapa.org/en/nevibecode/skills/bobmatnyc-dispatching-parallel-agents.md) — skill · 314 installs
- [parallel-task](https://affpapa.org/en/nevibecode/skills/parallel-task.md) — skill · 1200 installs
- [parallel-task-spark](https://affpapa.org/en/nevibecode/skills/parallel-task-spark.md) — skill · 1100 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md