OpenAI API pricing in 2026: GPT-5.6 family and beyond
OpenAI bills per million tokens with separate input, cached-input and output rates. The August 2026 lineup centers on the GPT-5.6 family — Sol (flagship), Terra (balanced) and Luna (fast/cheap) — with the GPT-5.4 family and GPT-4o still available for compatibility.
| Model | Input | Cached input | Output | Context |
|---|---|---|---|---|
| GPT-5.6 Sol | $5.00 | $0.50 | $30.00 | 1M |
| GPT-5.6 Terra | $2.00 | $0.20 | $12.00 | 1M |
| GPT-5.6 Luna | $0.20 | $0.02 | $1.20 | 1M |
| GPT-5.5 Pro | $30.00 | — | $180.00 | 1M |
| GPT-5.4 | $2.50 | $0.25 | $15.00 | 1M |
| GPT-5.4 mini | $0.75 | $0.075 | $4.50 | 400K |
| GPT-5.4 nano | $0.20 | $0.02 | $1.25 | 400K |
| GPT-4o | $2.50 | $1.25 | $10.00 | 128K |
| GPT-4o mini | $0.15 | $0.075 | $0.60 | 128K |
Cached input bills at ~10% of the standard input rate on GPT-5.x models. The Batch API takes 50% off both input and output on supported models.
Which GPT model is worth paying for?
Terra is the production workhorse — GPT-5.6 quality at $2/$12 covers most chat, RAG and summarization loads. Luna at $0.20/$1.20 is one of the best price-performance deals on the market for classification, extraction and short generation. Sol earns its 2.5× premium on hard reasoning; GPT-5.5 Pro at $30/$180 is a specialist tool — at 6× Sol's output price, reserve it for tasks where nothing else passes your evals.
The bill is mostly output tokens
Output costs 6× input across the GPT-5.6 family, so trimming max_tokens and verbose formatting usually saves more than switching models. Combine with automatic prompt caching (repeated prefixes bill at 10%) and the Batch API (−50%) before paying for a bigger tier — worked examples in our caching guide.
→ What would YOUR workload cost?Enter your token mix once — compare all 28 models side by side, with caching and batch discounts.