Skip to content

OpenAI Setup

Overview

Direct access to OpenAI's latest models (GPT-4, GPT-4 Turbo, o1) via the OpenAI API.

Prerequisites

  1. OpenAI Account with API access (https://platform.openai.com/signup)
  2. API Key from https://platform.openai.com/api-keys
  3. Python 3.8+
  4. Billing enabled on your OpenAI account

Step 1: Create an API Key

  1. Go to https://platform.openai.com/api-keys
  2. Click Create new secret key
  3. Copy the key (you won't see it again)
  4. Restrict key permissions if needed (recommended for production)

Step 2: Configure Environment

Edit your .env file:

# OpenAI
OPENAI_API_KEY=sk-...your-key...
OPENAI_MODEL=gpt-4-turbo  # or gpt-4, gpt-3.5-turbo, o1 (preview)

Step 3: Available Models (May 2026)

Latest: - gpt-4-turbo – Best reasoning, vision support - gpt-4o – Optimized for speed/cost - o1-preview – Reasoning-focused (limited availability)

Standard: - gpt-4 – Original GPT-4 - gpt-3.5-turbo – Fast, cheap

Context windows: - GPT-4 Turbo: 128K tokens - GPT-4o: 128K tokens - o1-preview: 128K tokens (reasoning steps don't count toward limit)

Full model list

Step 4: Run an Experiment

# Activate venv and install deps
source .venv-exp/bin/activate
pip install -r requirements-experiments.txt

# Run preset sweep
python3 scripts/cloud/openai_sweep.py --preset summarization --trials 3

# Analyze results
python3 scripts/analyze_results.py --input runs/latest.jsonl

Cost Estimates

OpenAI uses per-token pricing. Typical costs per 1000 API calls:

  • GPT-3.5 Turbo: ~$0.01-0.03
  • GPT-4 Turbo: ~$0.30-0.90
  • GPT-4o: ~$0.15-0.45
  • o1-preview: ~$3.00-15.00

Current pricing

Monitoring Usage

Check your usage:

# In your account at https://platform.openai.com/usage
# Or via CLI:
curl https://api.openai.com/v1/usage/requests \
  -H "Authorization: Bearer $OPENAI_API_KEY"

Set spending limits:

Go to https://platform.openai.com/account/billing/limits and set: - Hard limit (when to stop) - Soft limit (warn when approaching)

Troubleshooting

"Invalid API Key" error: Verify your key at https://platform.openai.com/api-keys. Delete and create a new one if needed.

"Rate limit exceeded" error: OpenAI has per-minute and per-day limits. Check your account tier and wait before retrying.

"Model not found" error: Verify the model name at https://platform.openai.com/docs/models. Some models (like o1) are in limited preview.

"Quota exceeded" error: Increase your spending limit in the console or wait for monthly reset.

Next: Run your first experiment