vs
0:2
overall capability wins (0 ties across 2 scored). o3-pro leads overall.
- Pricing3:0
- Speed2:0
- Specs0:2
Verdict
o3-pro offers ~1.6x larger context (200K vs 128K); o3-pro allows ~6.1x more max output (100K vs 16K); GPT-4o is ~8.0x cheaper on a blended token basis than o3-pro.
- o3-pro offers ~1.6x larger context (200K vs 128K)
- o3-pro allows ~6.1x more max output (100K vs 16K)
- GPT-4o is ~8.0x cheaper on a blended token basis than o3-pro
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- GPT-4o
- $5
- o3-pro
- $40
GPT-4o is about 8.0x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
OpenAI
OpenAI
Family
GPT-4 / o-series
GPT-4 / o-series
License
Proprietary
Proprietary
Open weights
No
No
Release
May 13, 2024
Jun 10, 2025
Knowledge cutoff
2023-10
-
API / provider
OpenAI
OpenAI
Modalities
text, image → text
text, image → text
Specs
Context window
128K
200K
Max output
16K
100K
Parameters
—
—
Pricing
Input $/1M
$2.50
$20
Output $/1M
$10
$80
Blended $/1M (3∶1)
$4.38
$35
Speed
tok/s
110
40
TTFT (s)
0.3
2.5
Reasoning
MMLU-Pro
—
—
GPQA Diamond
—
84.5
Humanity's Last Exam
—
—
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
—
—
AA-LCR v1.1
—
—
CritPt
—
—
MMMU-Pro
—
—
IFBench
—
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
—
—
Aider Polyglot
—
84.9
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
—
—
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
—
—
| Metric | GPT-4o | o3-pro | Delta |
|---|---|---|---|
| Identity | |||
| Organization | OpenAI | OpenAI | — |
| Family | GPT-4 / o-series | GPT-4 / o-series | — |
| License | Proprietary | Proprietary | — |
| Open weights | No | No | — |
| Release | May 13, 2024 | Jun 10, 2025 | — |
| Knowledge cutoff | 2023-10 | - | — |
| API / provider | OpenAI | OpenAI | — |
| Modalities | text, image → text | text, image → text | — |
| Specs | |||
| Context window | 128K | 200K | B +72K |
| Max output | 16K | 100K | B +84K |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $2.50 | $20 | A +$17.50 |
| Output $/1M | $10 | $80 | A +$70 |
| Blended $/1M (3∶1) | $4.38 | $35 | A +$30.63 |
| Speed | |||
| tok/s | 110 | 40 | A +70 |
| TTFT (s) | 0.3 | 2.5 | A +2.20 |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | — | 84.5 | — |
| Humanity's Last Exam | — | — | — |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | — | — | — |
| AA-LCR v1.1 | — | — | — |
| CritPt | — | — | — |
| MMMU-Pro | — | — | — |
| IFBench | — | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | — | — | — |
| Aider Polyglot | — | 84.9 | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | — | — | — |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | — | — | — |
GPT-4o
- Context
- 128K / 16K out
- Parameters
- —
- Price
- $2.50 / $10
- Speed
- 110 tok/s · 0.3s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Omnimodal GPT-4 class model; retained for multimodal latency and ecosystem compatibility.
o3-pro
- Context
- 200K / 100K out
- Parameters
- —
- Price
- $20 / $80
- Speed
- 40 tok/s · 2.5s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Maximum o3 reasoning tier for hardest STEM and multi-step problems at premium $20/$80 rates.
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.