vs
11:4
overall capability wins (3 ties across 18 scored). GPT-5.5 leads overall.
- Reasoning6:0
- Coding3:3
- Arena1:0
- Agents0:1
- Tool use1:0
- Composite indices1:1
- Pricing0:3
- Speed0:2
Verdict
GPT-5.5 edges reasoning & knowledge; GPT-5.5 ranks higher on LMArena (+24 Elo); GPT-5.6 Luna is ~25.0x cheaper on a blended token basis than GPT-5.5.
- GPT-5.5 edges reasoning & knowledge
- GPT-5.5 ranks higher on LMArena (+24 Elo)
- GPT-5.6 Luna is ~25.0x cheaper on a blended token basis than GPT-5.5
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- GPT-5.5
- $12.50
- GPT-5.6 Luna
- $0.50
GPT-5.6 Luna is about 25.0x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
OpenAI
OpenAI
Family
GPT-5.x
GPT-5.x
License
Proprietary
Proprietary
Open weights
No
No
Release
Apr 23, 2026
Jul 9, 2026
Knowledge cutoff
-
-
API / provider
OpenAI
OpenAI
Modalities
text, image → text
text, image → text
Specs
Context window
1.1M
1.1M
Max output
128K
128K
Parameters
—
—
Pricing
Input $/1M
$5
$0.20
Output $/1M
$30
$1.20
Blended $/1M (3∶1)
$11.25
$0.45
Speed
tok/s
70
195
TTFT (s)
0.5
0.2
Reasoning
MMLU-Pro
—
—
GPQA Diamond
93.5
91.1
Humanity's Last Exam
45.8
39.5
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
58.0
42.7
AA-LCR v1.1
84.3
83.7
CritPt
27.1
20.6
MMMU-Pro
79.9
78.6
IFBench
75.9
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
58.6
62.7
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
84.3
80.9
Aider Polyglot
—
—
Terminal-Bench 3
—
14.3
BigCodeBench
—
—
SciCode
55.8
53.6
CursorBench
58.4
61.1
SWE-Rebench
62.7
43.6
NL2Repo-Bench
—
—
DeepSWE
67.0
67.0
WebDev Arena
1,510
1,519
Terminal-Bench 4.0
—
17.3
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,476
1,453
| Metric | GPT-5.5 | GPT-5.6 Luna | Delta |
|---|---|---|---|
| Identity | |||
| Organization | OpenAI | OpenAI | — |
| Family | GPT-5.x | GPT-5.x | — |
| License | Proprietary | Proprietary | — |
| Open weights | No | No | — |
| Release | Apr 23, 2026 | Jul 9, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | OpenAI | OpenAI | — |
| Modalities | text, image → text | text, image → text | — |
| Specs | |||
| Context window | 1.1M | 1.1M | tie |
| Max output | 128K | 128K | tie |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $5 | $0.20 | B +$4.80 |
| Output $/1M | $30 | $1.20 | B +$28.80 |
| Blended $/1M (3∶1) | $11.25 | $0.45 | B +$10.80 |
| Speed | |||
| tok/s | 70 | 195 | B +125 |
| TTFT (s) | 0.5 | 0.2 | B +0.30 |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | 93.5 | 91.1 | A +2.4 pts |
| Humanity's Last Exam | 45.8 | 39.5 | A +6.3 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 58.0 | 42.7 | A +15.3 pts |
| AA-LCR v1.1 | 84.3 | 83.7 | A +0.6 pts |
| CritPt | 27.1 | 20.6 | A +6.5 pts |
| MMMU-Pro | 79.9 | 78.6 | A +1.3 pts |
| IFBench | 75.9 | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | 58.6 | 62.7 | B +4.1 pts |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | 84.3 | 80.9 | A +3.4 pts |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | 14.3 | — |
| BigCodeBench | — | — | — |
| SciCode | 55.8 | 53.6 | A +2.2 pts |
| CursorBench | 58.4 | 61.1 | B +2.7 pts |
| SWE-Rebench | 62.7 | 43.6 | A +19.1 pts |
| NL2Repo-Bench | — | — | — |
| DeepSWE | 67.0 | 67.0 | tie |
| WebDev Arena | 1,510 | 1,519 | B +9 Elo |
| Terminal-Bench 4.0 | — | 17.3 | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,476 | 1,453 | A +24 Elo |
GPT-5.5
- Context
- 1.1M / 128K out
- Parameters
- —
- Price
- $5 / $30
- Speed
- 70 tok/s · 0.5s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Prior OpenAI flagship with strong coding and long context; still widely used alongside the newer 5.6 family.
GPT-5.6 Luna
- Context
- 1.1M / 128K out
- Parameters
- —
- Price
- $0.20 / $1.20
- Speed
- 195 tok/s · 0.2s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Fastest, cheapest GPT-5.6 tier for high-volume chat, classification, and light coding; July 30, 2026 API cut dropped Standard rates 80% to $0.20/$1.20.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.