vs
0:0
overall capability wins (2 ties across 2 scored). GPT-5 Nano leads overall.
- Pricing0:3
Verdict
GPT-5 Nano is ~204.5x cheaper on a blended token basis than GPT-5.6 Cyber.
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- GPT-5.6 Cyber
- $31.25
- GPT-5 Nano
- $0.15
GPT-5 Nano is about 208.3x 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
Aug 10, 2026
Aug 7, 2025
Knowledge cutoff
2026-02-16
-
API / provider
OpenAI
OpenAI
Modalities
text, image → text
text, image → text
Specs
Context window
400K
400K
Max output
128K
128K
Parameters
—
—
Pricing
Input $/1M
$12.50
$0.05
Output $/1M
$75
$0.40
Blended $/1M (3∶1)
$28.13
$0.14
Speed
tok/s
-
200
TTFT (s)
-
0.15
Reasoning
MMLU-Pro
—
—
GPQA Diamond
—
67.6
Humanity's Last Exam
—
9.5
AIME 2025
—
85.2
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
—
19.1
AA-LCR v1.1
—
45.0
CritPt
—
—
MMMU-Pro
—
61.0
IFBench
—
67.6
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
54.7
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
—
—
Aider Polyglot
—
48.4
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
—
36.6
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
—
1,338
| Metric | GPT-5.6 Cyber | GPT-5 Nano | Delta |
|---|---|---|---|
| Identity | |||
| Organization | OpenAI | OpenAI | — |
| Family | GPT-5.x | GPT-5.x | — |
| License | Proprietary | Proprietary | — |
| Open weights | No | No | — |
| Release | Aug 10, 2026 | Aug 7, 2025 | — |
| Knowledge cutoff | 2026-02-16 | - | — |
| API / provider | OpenAI | OpenAI | — |
| Modalities | text, image → text | text, image → text | — |
| Specs | |||
| Context window | 400K | 400K | tie |
| Max output | 128K | 128K | tie |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $12.50 | $0.05 | B +$12.45 |
| Output $/1M | $75 | $0.40 | B +$74.60 |
| Blended $/1M (3∶1) | $28.13 | $0.14 | B +$27.99 |
| Speed | |||
| tok/s | - | 200 | — |
| TTFT (s) | - | 0.15 | — |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | — | 67.6 | — |
| Humanity's Last Exam | — | 9.5 | — |
| AIME 2025 | — | 85.2 | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | — | 19.1 | — |
| AA-LCR v1.1 | — | 45.0 | — |
| CritPt | — | — | — |
| MMMU-Pro | — | 61.0 | — |
| IFBench | — | 67.6 | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | 54.7 | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | — | — | — |
| Aider Polyglot | — | 48.4 | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | — | 36.6 | — |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | — | 1,338 | — |
GPT-5.6 Cyber
- Context
- 400K / 128K out
- Parameters
- —
- Price
- $12.50 / $75
- Speed
- —
- Modalities
- text, image → text
- License
- Proprietary
OpenAI's most capable and most permissive cybersecurity model (Daybreak program) for authorized vulnerability research and exploit validation — a GPT-5.6 Sol derivative at $12.5/$75; gated behind identity verification and legal declarations.
GPT-5 Nano
- Context
- 400K / 128K out
- Parameters
- —
- Price
- $0.05 / $0.40
- Speed
- 200 tok/s · 0.15s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Ultra-cheap OpenAI tier for classification, extraction, and simple assistants.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.
These two models have no overlapping published benchmarks in our dataset. Compare specs and pricing instead.