# Paper Lantern

Connects coding agents to 2M+ computer science research papers to evaluate algorithmic tradeoffs, benchmarks, and architectural decisions.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI
- Author: paperlantern-ai
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/paperlantern-ai/paperlantern-cli
- Updated: 2026-04-21
- HTML: https://affpapa.org/en/nevibecode/mcp/paper-lantern

## What it does

The server equips coding assistants with academic CS literature. Whenever an assistant faces an architectural dilemma, it queries research papers for candidate trade-offs and benchmark results. The integration stays inactive during basic syntax lookup or standard code debugging.

Five dedicated tools handle the research workflow: `explore_approaches`, `deep_dive`, `compare_approaches`, `check_feasibility`, and `give_feedback`. The feasibility check evaluates project boundaries to return an explicit GO, PROTOTYPE, or RECONSIDER verdict.

Setup via the CLI opens a desktop browser for OAuth verification. Because of this browser step, running the automated wizard on a headless machine will fail without manual endpoint configuration.

## Requirements
- Remote server: `https://mcp.paperlantern.ai/chat/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http code https://mcp.paperlantern.ai/chat/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add paper-lantern --url https://mcp.paperlantern.ai/chat/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "paper-lantern": {
            "url": "https://mcp.paperlantern.ai/chat/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "paper-lantern": {
            "type": "http",
            "url": "https://mcp.paperlantern.ai/chat/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http paper-lantern https://mcp.paperlantern.ai/chat/mcp
```

## When to use it
- evaluate trade-offs between competing algorithm families
- verify architectural feasibility before writing production code
- analyze hyperparameters and failure modes of a machine learning method
- compare academic benchmark results side by side

## Example prompts
- Compare candidate algorithms for real-time deduplication under strict memory constraints
- Check feasibility of running an approximate nearest neighbor index within 256MB RAM
- Do a deep dive on modern cache eviction policies and their known failure modes

## Similar
- [Zotero MCP](https://affpapa.org/en/nevibecode/mcp/zotero-mcp.md) — MCP server · ★ 5585
- [arxiv-mcp-server](https://affpapa.org/en/nevibecode/mcp/arxiv-mcp-server.md) — MCP server · ★ 3235
- [PaperBanana](https://affpapa.org/en/nevibecode/mcp/paperbanana.md) — MCP server · ★ 2418
- [MATLAB MCP Server](https://affpapa.org/en/nevibecode/mcp/matlab-mcp-server.md) — MCP server · ★ 1723
- [Hex Research](https://affpapa.org/en/nevibecode/mcp/hex-research.md) — MCP server · ★ 584
- [DocsAgent — Zotero, Obsidian & Apple Notes MCP Server](https://affpapa.org/en/nevibecode/mcp/docsagent-zotero-obsidian-apple-notes-mcp-server.md) — MCP server · ★ 578

---
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