# Deep Research Openai

> Run OpenAI deep research models through the Responses API with explicit effort, model, cost, and output controls. Use when launching citation-backed OpenAI research runs, smoke-testing OpenAI provider behavior, or capturing reproducible benchmark output.

- Skill: `closedloop-technologies/deep-research-openai` (Agent Skill, multi-file: 5 files)
- Install (CLI): `npx skillmds@latest add closedloop-technologies/deep-research-openai`
- Raw SKILL.md: https://api.skillmd.com/api/skills/closedloop-technologies/deep-research-openai/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: closedloop-technologies (https://skillmd.com/u/closedloop-technologies)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/closedloop-technologies/deep-research-openai

---


# OpenAI Deep Research

Use this skill for autonomous, citation-backed research through OpenAI's Deep
Research models. Prefer `o4-mini-deep-research` for benchmark calls because the
repo target is less than `$1` per task.

## Environment

```bash
op run --env-file .env.climbhill -- python .agents/skills/deep-research-openai/scripts/run_deep_research.py \
  --prompt "deepresearch the deepresearchers" \
  --model o4-mini-deep-research \
  --effort low
```

Required 1Password-backed variable:

- `OPENAI_API_KEY`

## Cost Controls

- Use `o4-mini-deep-research` before `o3-deep-research`.
- Use `--effort low` for benchmark smoke tasks.
- Keep prompts narrow and request concise reports.
- Save raw output, then normalize it with `deep-research-okf-normalize`.

## Current Source Links

- https://platform.openai.com/docs
- https://platform.openai.com/docs/guides/deep-research