Become a 10x Brand Engineer with Cuppa AI See Who's Hiring ↗
Getting StartedOpenAI API

OpenAI API

OpenAI API Setup

Connect your OpenAI API key to Cuppa to generate content using GPT models. This is required for all Cuppa users.

Why OpenAI?

OpenAI’s GPT models power many of Cuppa’s features:

  • Article generation (single and bulk)
  • Outline creation
  • Metadata generation
  • Content optimization

Even if you prefer other models (Claude, Grok), OpenAI is required for internal features.

Available Models

ModelIDContextBest For
GPT-6 Astragpt-6-astra1.05MFlagship. Long-horizon writing, professional work, hard briefs
GPT-5.6 Solgpt-5.6-sol1.05MPrevious flagship for coding, knowledge work, and science
GPT-5.6 Terragpt-5.6-terra1.05MBalanced everyday articles at lower cost than Sol
GPT-5.6 Lunagpt-5.6-luna1.05MFastest, cheapest GPT-5.6 tier for high-volume drafts
GPT-5.3 Instantgpt-5.3-chat-latest128kConversational writing with fewer hallucinations
GPT-5 Minigpt-5-mini128kDefault for high-volume API / Slack generation

GPT-6 Astra is marked NEW in every text model picker. List the live catalog with GET /v1/meta/models or MCP list_models.

Recommendation: Use Luna or Mini for volume. Use Terra for daily articles. Use Astra when the piece has to be right.

GPT Image 2.5

The same OpenAI API key unlocks GPT Image 2.5 Sunburst and GPT Image 2.5 Flare. Both are marked NEW in every image picker: Creative Studio, articles, Content Planner, local clusters, Agent Chat, and GET /v1/meta/image_models.

ModelIDBest For
GPT Image 2.5 Sunburstgpt-image-2.5-sunburstFlagship. Edits, precision, highest fidelity
GPT Image 2.5 Flaregpt-image-2.5-flareFast everyday generation
GPT Image 2gpt-image-2Previous flagship
GPT Image 1.5gpt-image-1.5Existing workflows
GPT Image 1gpt-image-1Default in some article / Slack flows

Quality options on 2.5: low, medium, high, xhigh, max, auto. Token rates match GPT Image 2 (5textin/5 text in / 8 image in / $30 image out per 1M tokens).

Via API / CLI / MCP, pass model: "gpt-image-2.5-sunburst" or model: "gpt-image-2.5-flare". See Image Generation.

Step 1: Create an OpenAI API Account

  1. Go to platform.openai.com
  2. Click Sign Up (or Log In)
  3. Complete account verification

Note: This is different from ChatGPT. You need a separate API account.

Step 2: Add Payment & Credits

OpenAI API requires prepaid credits:

  1. Go to Settings → Billing
  2. Click Add payment method
  3. Add a credit card
  4. Click Add to credit balance
  5. Start with $10-20 for testing

API Tiers

OpenAI has usage tiers that affect rate limits:

TierRequirementRate Limits
FreeNew accountVery limited (won’t work well)
Tier 1$5+ spentGood for single generation
Tier 2$50+ spentRecommended for Cuppa
Tier 3$100+ spentBest for heavy bulk generation

For bulk generation (10+ articles), Tier 2 is recommended.

Step 3: Generate Your API Key

  1. Go to platform.openai.com/api-keys
  2. Click Create new secret key
  3. Name it “Cuppa” (optional)
  4. Click Create secret key
  5. Copy the key immediately - you won’t see it again!

Your key looks like: sk-proj-xxxxxxxxxxxxxxxxxxxx

Step 4: Add Key to Cuppa

  1. Log in to app.cuppa.ai
  2. Go to Team Settings → API Keys
  3. Find the OpenAI section
  4. Paste your API key
  5. Click Save

Cuppa will validate your key before saving.

Verify It’s Working

Test your connection:

  1. Go to Create Content → Single Generation
  2. Enter any keyword (e.g., “test article”)
  3. Select a GPT model
  4. Click Generate

If it works, you’re all set!

Multiple API Keys (Business+)

On Business plans and higher, you can add multiple OpenAI keys for:

  • Faster bulk generation - parallel API calls
  • Higher rate limits - spread requests across keys
  • Redundancy - if one key hits limits, others continue

To add multiple keys:

  1. Go to Team Settings → API Keys
  2. Click Add Another Key
  3. Paste your additional key

Protected Models

GPT-6 Astra is rolling out on OpenAI’s side (Trusted Access first, then Plus / Pro / Business / Enterprise and API). If the model appears in Cuppa but OpenAI returns an access error, the key’s organization is not enabled for Astra yet.

Some newer OpenAI models also require organization verification:

  1. Go to platform.openai.com/settings/organization
  2. Complete Organization Verification
  3. Wait for approval (usually 1-2 business days)

If a model isn’t working, check if it requires verification.

Cost Management

Monitoring Usage

Track your API costs:

  1. Go to platform.openai.com/usage
  2. View daily and monthly spending
  3. Set usage limits to prevent surprises

Setting Limits

Protect against unexpected costs:

  1. Go to Settings → Limits
  2. Set a Monthly budget (e.g., $50)
  3. Set a Notification threshold (e.g., 80%)

Troubleshooting

”Invalid API key”

  • Check you copied the full key (starts with sk-)
  • Verify the key hasn’t been deleted in OpenAI
  • Create a new key if unsure

”Insufficient quota” or “Rate limit”

  • Add more credits to your OpenAI account
  • Upgrade to Tier 2 ($50+ spent)
  • Wait a minute and try again
  • Use multiple API keys (Business+)

“Model not found”

  • Some models require organization verification
  • Check the model name is correct
  • Ensure your account has access to that model

”Free credits don’t work”

OpenAI’s free trial credits don’t work with Cuppa. Add a payment method and prepay credits.

Generation takes too long

  • Check status.openai.com for outages
  • Try a faster model (GPT-4o-mini)
  • Reduce article length/complexity

FAQ

Is OpenAI API the same as ChatGPT Plus?

No. ChatGPT Plus ($20/month) is for the chat interface. The API is separate with pay-per-use pricing.

Why do I need OpenAI even with Claude?

Cuppa uses OpenAI for certain internal features (outline generation, SEO metadata). Your main article content can use any model.

How do I check my OpenAI tier?

Go to platform.openai.com/settings/organization/limits to see your current tier.

Can I use my employer’s OpenAI account?

Yes, if you have access to create API keys. Just copy the key to Cuppa.

What if I run out of credits mid-generation?

The generation will fail. Top up your credits and retry - partial progress is usually saved.

Are my API keys secure?

Yes. Keys are encrypted at rest and never shared. You can revoke a key anytime from OpenAI.

Related Docs

  • Anthropic API Setup - Add Claude models
  • Perplexity API Setup - Enable auto-research
  • Team Settings - Manage your API keys
Quick Start Guide and TourAnthropic API