Cheaper overall
gpt-5.6-luna
90% lower blended rate
Higher benchmark scores
gpt-5.6-terra
leads on 8 of 11 figures
Cheaper cached input
gpt-5.6-luna
$0.02 vs $0.20, 90% less
Where each one wins
gpt-5.6-luna
- Cheaper input tokens $0.20 vs $2.00, 90% less
- Cheaper output tokens $1.20 vs $12.00, 90% less
- Cheaper cached input $0.02 vs $0.20, 90% less
- Cheaper on all four workloads
- Higher long-context reasoning (AA-LCR) 83.7 vs 83
- Higher real-world task performance (GDPval) 47.2 vs 46.6
- Higher answer reliability (Omniscience Non-Hallucination) 25 vs 12.1
gpt-5.6-terra
- Higher overall intelligence (Intelligence Index) 42.1 vs 37.3
- Higher coding ability (Coding Index) 76.7 vs 71.4
- Higher multi-step tool use (Agentic Index) 43.2 vs 42.1
- Ahead on 5 more benchmark figures
What four workloads cost
One run, and the same run a thousand times.
gpt-5.6-luna is cheaper on all four workloads, by much the same margin each time — about 10 times, from $0.0008 against $0.0080 on a chat turn to $0.1654 against $1.65 on a long document.
| Workload | gpt-5.6-luna | gpt-5.6-terra | ×1,000 | Difference |
|---|---|---|---|---|
| Chat turn 1,000 in / 500 out | $0.0008 | $0.0080 | $0.8000 / $8.00 | gpt-5.6-luna is 90% cheaper |
| RAG answer 10,000 in / 800 out | $0.0030 | $0.0296 | $2.96 / $29.60 | gpt-5.6-luna is 90% cheaper |
| Agent loop 32,000 in / 700 out × 20 calls | $0.0368 | $0.3680 | $36.80 / $368.00 | gpt-5.6-luna is 90% cheaper |
| Long document 400,000 in / 3,000 out | $0.1654 | $1.65 | $165.40 / $1,654.00 | gpt-5.6-luna is 90% cheaper |
Price per 1M tokens
Above 272,000 input tokens the rates change, and the new rate applies to the whole request, not just the tokens past that point: gpt-5.6-luna's input goes from $0.2000 to $0.4000, and gpt-5.6-terra's input goes from $2.00 to $4.00. It does not change which of the two is cheaper.
| Context band | Price | gpt-5.6-luna | gpt-5.6-terra | Difference |
|---|---|---|---|---|
| Prompts up to 272,000 tokens | Input | $0.20 | $2.00 | gpt-5.6-luna is 90% cheaper |
| Prompts up to 272,000 tokens | Cached input | $0.02 | $0.20 | gpt-5.6-luna is 90% cheaper |
| Prompts up to 272,000 tokens | Output | $1.20 | $12.00 | gpt-5.6-luna is 90% cheaper |
| Prompts over 272,000 tokens | Input | $0.40 | $4.00 | gpt-5.6-luna is 90% cheaper |
| Prompts over 272,000 tokens | Cached input | $0.04 | $0.40 | gpt-5.6-luna is 90% cheaper |
| Prompts over 272,000 tokens | Output | $1.80 | $18.00 | gpt-5.6-luna is 90% cheaper |
Blended: gpt-5.6-luna $0.45, gpt-5.6-terra $4.50 per 1M — one rate at 3:1 input to output, for comparing two models at a glance.
Both priced by OpenAI. Standard tier, pay-as-you-go.
Other pricing tiers
Both models offer Batch, Flex and Priority, worth up to 50% off their standard rates, and the cheaper model is the same one on every tier — so the tier you buy changes the bill rather than the choice.
Every tier either model sells is here, so this is the whole price sheet. A saving is measured against that model's own standard rate — how we read tiers.
| Tier | gpt-5.6-luna | gpt-5.6-terra | Cheaper |
|---|---|---|---|
| Batch | $0.10 in / $0.60 out 50% off Standard | $1.00 in / $6.00 out 50% off Standard | gpt-5.6-luna |
| Flex | $0.10 in / $0.60 out 50% off Standard | $1.00 in / $6.00 out 50% off Standard | gpt-5.6-luna |
| Priority | $0.40 in / $2.40 out | $4.00 in / $24.00 out | gpt-5.6-luna |
Benchmarks
They share 11 figures: gpt-5.6-luna leads on 3, gpt-5.6-terra on 8. The widest gap is 12.9 points, on answer reliability (Omniscience Non-Hallucination).
Overall intelligence
Intelligence Index — Overall intelligence across reasoning, knowledge & math evals
gpt-5.6-terra ahead by 4.8
Coding ability
Coding Index — Coding ability across software-engineering evals
gpt-5.6-terra ahead by 5.3
Multi-step tool use
Agentic Index — Tool use & multi-step agent task performance
gpt-5.6-terra ahead by 1.1
Graduate-level science
GPQA Diamond — Graduate-level questions in biology, chemistry & physics
gpt-5.6-terra ahead by 1.4
Expert-exam performance
Humanity's Last Exam — Expert-level questions across many academic domains
gpt-5.6-terra ahead by 3.4
Long-context reasoning
AA-LCR — Long-context reasoning across large inputs
gpt-5.6-luna ahead by 0.7
Scientific coding
SciCode — Research-level scientific coding tasks
gpt-5.6-terra ahead by 1.4
Physics research reasoning
CritPt — Unpublished physics research reasoning problems
gpt-5.6-terra ahead by 9.4
Real-world task performance
GDPval — Economically valuable, real-world knowledge work
gpt-5.6-luna ahead by 0.6
Factual accuracy
Omniscience Accuracy — Breadth of factual knowledge across domains
gpt-5.6-terra ahead by 4.1
Answer reliability
Omniscience Non-Hallucination — How reliably the model avoids fabricated answers
gpt-5.6-luna ahead by 12.9
Each figure is that model's best published run, with the effort level named beside it — where these scores come from.
Specifications
Both list the same capabilities, so nothing here separates them.
| Specification | gpt-5.6-luna | gpt-5.6-terra |
|---|---|---|
| Context window | 1,050,000 tokens | 1,050,000 tokens |
| Max output | 128,000 tokens | 128,000 tokens |
| Takes in | Text, image, file | Text, image, file |
| Puts out | Text | Text |
| Knowledge cutoff | 16 Feb 2026 | 16 Feb 2026 |
| Released | 9 Jul 2026 | 9 Jul 2026 |
| Status | Active | Active |
| Sold by | OpenAI, Perplexity | OpenAI, Perplexity |
| Tool use | Yes | Yes |
| Structured outputs | Yes | Yes |
| Web search | Not listed | Not listed |
| Prompt caching | Yes | Yes |
| Code execution | Yes | Yes |
| Computer use | Not listed | Not listed |
| Open weights | Not listed | Not listed |
| Reasoning | Yes | Yes |
FAQs
- Is gpt-5.6-luna cheaper than gpt-5.6-terra?
- Yes. A 1,000-token prompt with a 500-token reply costs $0.0008 on gpt-5.6-luna and $0.0080 on gpt-5.6-terra, and the same model is cheaper on every workload on this page.
- How much do gpt-5.6-luna and gpt-5.6-terra cost per 1M tokens?
- gpt-5.6-luna costs $0.20 for input and $1.20 for output. gpt-5.6-terra costs $2.00 and $12.00. Standard pay-as-you-go rates, as of 30 Jul 2026.
- Which is better, gpt-5.6-luna or gpt-5.6-terra?
- On published benchmarks, gpt-5.6-terra. It leads on 8 of the 11 figures both models report, including overall intelligence (Intelligence Index), where it scores 42.1 against 37.3.
- Is Batch pricing cheaper for gpt-5.6-luna or gpt-5.6-terra?
- Batch input costs $0.10 on gpt-5.6-luna and $1.00 on gpt-5.6-terra, 50% below the standard rate on both.