OpenAI Models Landscape
Complete Reference & Cost Estimator
A comprehensive breakdown of every model in OpenAI’s current API portfolio — organized by tier, use case, and cost — with an interactive calculator to estimate your production spend before you build.
Portfolio Architecture
Five Layers. One Portfolio.
OpenAI’s lineup is now structured around clear tiers. Understanding which layer fits your workload is more valuable than knowing every model name.
Release Timeline
Cadence Has Accelerated
OpenAI shipped more model families in 18 months than in the prior 3 years combined.
Key Strategic Themes
What Changed vs 2024
Shortlist Tier
Models to Actively Evaluate
These are the models Cohort 10 should consider first. Each card shows pricing, context, and best-fit role.
Side-by-Side
Top-16 Models Compared
Sorted by recommended tier. Prices are standard list in USD per 1M tokens unless stated. Input cached = prompt-cache discount price.
| Model | Best Role | Context | Input $/1M | Cached $/1M | Output $/1M | Latency |
|---|
Embedding & Moderation
Utility Models
| Model | Role | Price $/1M | Notes |
|---|---|---|---|
| text-embedding-3-small | Default retrieval/search | $0.02 | Best price/performance default |
| text-embedding-3-large | High-recall / multilingual | $0.13 | Better recall on hard tasks |
| text-embedding-ada-002 | Legacy compat only | $0.10 | Prefer 3-small for new builds |
| omni-moderation-latest | Text + image safety filter | FREE | Use as pre/post-filter on all endpoints |
Interactive Tool
Estimate Your Monthly Cost
Enter your workload parameters below. The calculator shows side-by-side cost across the most common models so you can make the right build vs. quality trade-off before writing a single line of code.
💰 OpenAI API Cost Estimator
Based on official list pricing as of May 2026. Does not include Batch/Flex discounts.
Quick presets →
Monthly cost by model →
Workload → Model
Task-to-Model Reference
Match your use case to the right model before you build. Primary = best quality. Alternative = lower cost or latency. Hover any row for details.
Decision Tree
How to Pick the Right Model
Follow this logic for every new project. When in doubt, start with GPT-5.4 mini and escalate only when needed.
Migration Paths
Upgrading Older Fleets
If you’re moving off legacy models, use these direct upgrade paths.
Architecture Pattern
Recommended Router Architecture
Cost Optimization