Turn your Idea to Paper by one Click

🚀 Launch Project ⚡ Deploy Locally

Offload the Labor. Steer the Science.

About Idea2Paper

From Idea to Paper, Automatically

Idea2Paper is an idea-to-paper research automation service. Give it a research idea and a target venue — Idea2Paper handles literature review, experiment design, code execution, figure generation, LaTeX writing, and iterative revision. Six specialized AI agents collaborate through a structured pipeline, while you stay in control at every checkpoint via a web dashboard or Telegram.

Demo

See Idea2Paper in Action

A walkthrough of Idea2Paper taking a research idea from proposal to publication-ready paper.

Self-host

Install in One Line

Linux or macOS. Builds the conda envs, installs Idea2Paper + Claude Code + Gemini CLIs, runs the dashboard as a systemd --user service, and prints a one-time magic-link URL for your local sign-in. No SMTP, no Google OAuth needed.

curl -fsSL https://idea2paper.org/install.sh | bash
--no-webapp skip the local dashboard service --no-research skip optional research extras --prefix DIR install to a custom directory --dry-run preview without changes

The script asks for your Gemini API key, Claude OAuth token, and login email. After install: click the printed magic link → you're in the dashboard at http://localhost:9527. Full docs at doc.html · script at install.sh.

Results

Example Papers

Real papers generated end-to-end by Idea2Paper.

Budget-Constrained Multi-Modal Research Synthesis Template: EuroMLSys

Budget-Constrained Multi-Modal Research Synthesis via Iterative-Deepening Agentic Search

Read Paper
HeteroServe Paper Template: ICML

HeteroServe: Capability-Weighted Batch Scheduling for Heterogeneous GPU Clusters in LLM Inference

Read Paper
TierKV Paper Template: NeurIPS

TierKV: Prefetch-Aware Memory Tiering for KV Cache in LLM Serving

Read Paper

Venue Intelligence

Supported Venues

NeurIPS ICML ICLR AAAI ACL IEEE ACM SIGPLAN LNCS USENIX CVPR ECCV NeurIPS ICML ICLR AAAI ACL IEEE ACM SIGPLAN LNCS USENIX CVPR ECCV NeurIPS ICML ICLR AAAI ACL IEEE ACM SIGPLAN LNCS USENIX CVPR ECCV

Why Idea2Paper

Full Automation, Full Control

Other Tools
  • Autonomous Drift
    Runs unsupervised; no way to course-correct mid-run.
  • Broken Formatting
    Inconsistent layouts and broken LaTeX need manual cleanup.
  • Hallucinated Citations
    LLMs fabricate plausible-looking references to nonexistent papers.
  • Low-Quality Figures
    Default styles, wrong sizes, no awareness of page constraints.
  • Complex Setup
    Requires manual configuration, environment tuning, and dependency wrangling before first run.
Idea2Paper
  • Human-in-the-Loop
    Pauses at key decisions; steer via Telegram or web anytime.
  • Publication-Ready Layout
    Hard-coded LaTeX + venue templates (NeurIPS, ACL, IEEE…).
  • Verified References
    Every citation checked against DBLP — no fake references.
  • Precision-Controlled Figures
    Paper Banana + venue-aware canvas, column widths, and fonts.
  • Click to Start
    One-click launch from the web UI — zero configuration needed.

How It Works

Idea2Paper Pipeline

Research, develop, and review — iterating until the paper meets your quality target.

1

Research Phase

Deep Research gathers background knowledge and literature survey.

2

Development Phase

Plan experiments, run on compute, analyze results, evaluate completeness, write initial draft.

3

Review Phase

Iterative paper improvement until the reviewer score hits the acceptance threshold.

Research Phase

⚙️ Setup → 📖 Analyze Proposal → 🔬 Deep Research → 🎯 Specialization → 📚 Bootstrap

Development Phase

Iterative
📋 Plan Experiments → 🧪 Run Experiments → 📊 Analyze Results → ✅ Evaluate → 📈 Generate Figures → ✍️ Write Draft → 📦 Deliver

Review Phase

Until score ≥ threshold
🔍 Review → 📋 Plan → ⚡ Execute

Meet the Team

Idea2Paper Agents

Six specialists collaborating through shared state and structured handoffs.

Researcher

Analyzes proposals, generates deep research queries, and bootstraps citations.

Reviewer

Evaluates papers page-by-page and scores across multiple dimensions.

Planner

Classifies issues, creates remediation tasks, and plans experiments.

Writer

Modifies LaTeX papers and Python plotting scripts for figures.

Experimenter

Runs real experiments, collects and analyzes results.

Coder

Writes production code, tests, and maintains code quality.