Historical model (legacy). The rates below are the last recorded rates, not a recommendation for a new deployment.
GPT-5.6 Luna API Pricing
GPT-5.6 Luna costs $0.20 per 1M input tokens and $1.20 per 1M output tokens. These rates were last verified on .
Moving off GPT-5.6 Luna?
Same token usage, 3:1 input:output, standard list rates · excludes cache, batch and fast mode. Not a quality or compatibility claim. Full calculator →
OpenAI small general text pricing history
USD · units shown per rate- Input: $0.10 / 1M tokensOutput: $0.50 / 1M tokensRate first observedLatestVerified 2026-11-12
- Input: $0.20 / 1M tokensOutput: $1.20 / 1M tokensRate first observedAvailable · older versionVerified 2026-11-12
1 recorded price changes
- Input: $0.20 / 1M tokens · Output: $1.20 / 1M tokens — observed Jul 31, 2026
- Input: $1.00 / 1M tokens · Output: $6.00 / 1M tokens — observed Jun 28, 2026 (first recorded)
- Input: $0.75 / 1M tokensOutput: $4.50 / 1M tokensRate first observedAvailable · older versionVerified 2026-11-12
- Input: $0.20 / 1M tokensOutput: $1.25 / 1M tokensRate first observedDeprecated · 2027-04-01Verified 2026-11-12
- Input: $0.25 / 1M tokensOutput: $2.00 / 1M tokensRate first observedDeprecated · 2026-12-11Verified 2026-11-12
- Input: $0.05 / 1M tokensOutput: $0.40 / 1M tokensRate first observedDeprecated · 2026-12-11Verified 2026-11-12
- Input: $0.40 / 1M tokensOutput: $1.60 / 1M tokensRate first observedDeprecated · 2027-01-14Verified 2026-11-12
- Input: $0.10 / 1M tokensOutput: $0.40 / 1M tokensRate first observedShut downVerified 2026-11-12
- Input: $0.15 / 1M tokensOutput: $0.60 / 1M tokensRate first observedDeprecated · 2027-01-14Verified 2026-11-12
Newest first. Release dates describe the model, not when today's rates took effect. Observations show when we recorded rates, not launch pricing. Image and text token meters are separate; no fixed image-token savings mix is assumed. How we verify
Published rates and billing units
| Billing dimension | USD list price |
|---|---|
| Input / 1M tokens | $0.20 |
| Cached input / 1M tokens | $0.02 |
| Cache write (30 minutes) / 1M tokens | $0.25 |
| Requests above 272,000 input tokens: Input / 1M tokens | $0.40 |
| Requests above 272,000 input tokens: Cached input / 1M tokens | $0.04 |
| Requests above 272,000 input tokens: Cache write (30 minutes) / 1M tokens | $0.50 |
| Requests above 272,000 input tokens: Output / 1M tokens | $1.80 |
| Output / 1M tokens | $1.20 |
| Batch: Input / 1M tokens | $0.10 |
| Batch: Cached input / 1M tokens | $0.01 |
| Batch: Output / 1M tokens | $0.60 |
| Batch: Requests above 272,000 input tokens: Input / 1M tokens | $0.20 |
| Batch: Requests above 272,000 input tokens: Cached input / 1M tokens | $0.02 |
| Batch: Requests above 272,000 input tokens: Output / 1M tokens | $0.90 |
Verified 2026-11-12 · Official source · Pricing methodology. Taxes, optional services and account-specific discounts are not included.
Cost examples
These examples use published list prices only. They exclude taxes, enterprise discounts, minimum charges, retries, batch discounts, and provider-specific billing rules.
1M input + 1M output
$2.20
A direct per-million comparison against every other text model.
100K input + 25K output
$0.05
A compact chat, summarization, or analysis workload.
10M input + 2M output
$7.60
A monthly production estimate for heavier RAG or agent traffic.
75% cached input + 250K output
$0.58
Cached input is about 90% cheaper than standard input at list price.
Calculator for GPT-5.6 Luna
Enter your own input, output, and cached-token assumptions. This calculator is preloaded with only GPT-5.6 Luna, so the result stays focused on this model.
Input cost
$0.40
Output cost
$1.80
Total
$2.20
Long-context rates apply because this request exceeds 272,000 input tokens.
Price history
input moved from $1.00 to $0.20 and output moved from $6.00 to $1.20 across 137 daily snapshots.
When to use GPT-5.6 Luna
Good fit
- code generation, debugging, and agentic development work where output quality affects engineering time
- high-volume routing, classification, extraction, summaries, and fallback traffic
- long-document and repository-scale prompts where context window matters
Be careful with
- blind migration without testing your own prompts, latency, and failure modes