Across 9 comparable benchmarks, o3-pro has better reported results on 9 and worse on 0.
Moderate evidence9 of 9 shared results countedHow we compare
Price basis: Standard rates. The price section below shows the applicable periods.
| Fact | o3-pro | GPT-4.1 |
|---|---|---|
| Context window | 200K tokens | 1M tokens |
| Input / 1M | $20.00 | $2.00 |
| Output / 1M | $80.00 | $8.00 |
| Blended (4:1) | $32.00 | $2.12 |
| Status | Deprecated | Deprecated |
| Released | June 10, 2025 | April 14, 2025 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 7 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
o3-pro ahead on 3 of 3 counted
o3-pro ahead on 2 of 2 counted
o3-pro ahead on 1 of 1 counted
o3-pro ahead on 1 of 1 counted
No shared comparable results for Abstract reasoning, Computer & tool use, Research & search, Multimodal, Professional work & productivity, Cybersecurity.
Published scores, with matching modes where available. Gaps show o3-pro relative to GPT-4.1. Reading the results
| Benchmark | o3-pro | GPT-4.1 | Gap |
|---|---|---|---|
| Aider-PolyglotAgentic Development · Accuracy | 84.9Thinking High · no tools | 52.4≠Normal · no tools | +32.5 pts#2 vs #22 of 47 |
| SWE-bench VerifiedRepository Engineering · Accuracy | 75.0Thinking High · no tools | 54.6≠Normal · no tools | +20.4 pts#37 vs #85 of 108 |
| WeirdML v2ML Engineering · Average accuracy across 17 tasks (%) | 58.2Thinking High · tools | 39.0≠Normal · tools | +19.2 pts#21 vs #35 of 39 |
| AIME 2024Mathematics · Accuracy | 93.0Normal · no tools | 48.1Normal · no tools | +44.9 pts#7 vs #44 of 54 |
| AIME2025Mathematics · Accuracy | 93.0Normal · no tools | 36.7Normal · no tools | +56.3 pts#29 vs #82 of 86 |
| GPQA DiamondScientific Reasoning · Accuracy | 84.0Normal · no tools | 66.3Normal · no tools | +17.7 pts#88 vs #152 of 200 |
| Fiction.liveBenchLong Reasoning · 16k-token score (%) | 97.2Thinking Medium · no tools | 63.9≠Normal · no tools | +33.3 pts#1 vs #17 of 28 |
| HLEKnowledge Exams · Accuracy | 21.0Normal · no tools | 5.4≠Reported best | +15.6 pts#101 vs #141 of 151 |
| ECICapability Indices · ECI score (capability index, higher is better) | 147Reported best | 137Reported best | 1.08×#59 vs #108 of 168 |
Agentic Development · Accuracy
Repository Engineering · Accuracy
ML Engineering · Average accuracy across 17 tasks (%)
Scientific Reasoning · Accuracy
Long Reasoning · 16k-token score (%)
A dash means no result is available.
| Benchmark | o3-pro | GPT-4.1 |
|---|---|---|
| FrontierMath | — | 5.5 |
| FrontierMath - Tier 4 | — | 0.0 |
| FrontierMath v2 | — | 6.0 |
| GSM8K | — | 95.9 |
| MATH-500 | — | 92.8 |
| MMLU | — | 90.2 |
| MMLU-Pro | — | 80.5 |
| IC SWE-Lancer (Diamond) | — | 14.4 |
| LiveCodeBench | — | 40.5 |
| SWE Manager Lancer (Diamond) | — | 35.1 |
| Terminal Bench Hard | — | 13.6 |
| MCP-Atlas | 44.5 | — |
| Creative Writing | — | 1420 |
| DROP | — | 89.2 |
| GeoBench ACW | — | 72.0 |
| MMMU-Pro | — | 61.2 |
| τ²-Bench | — | 54.7 |
| τ²-Bench - Telecom | — | 34.0 |
| ARC-AGI-1 | 59.3 | — |
| ARC-AGI-2 | 4.9 | — |
| SimpleBench | — | 27.0 |
| GAIA | — | 49.7 |
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
o3-pro costs1,409%morethan GPT-4.1 per million tokens — 15.1× the price.Using standard rates at this token mix.
o3-pro: Standard pricing. OpenAI official pricing (developers.openai.com/api/docs/pricing) verified 2026-11-13; standard tier; Deprecated; shutdown 2026-12-11 (replacement: gpt-6.1-sol pro)
| Price / 1M tokens | o3-pro | GPT-4.1 | Ratio |
|---|---|---|---|
| Input | $20.00 | $2.00 | 10.0× |
| Output | $80.00 | $8.00 | 10.0× |
| Cached input | — | $0.50 | — |
| Cache write | — | — | — |
| Blended | $32.00 | $2.12 | 15.1× |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
Reported cost per task, paired with the score from the same run. Settings and sources may differ; costs do not affect the verdict.
| Attribute | o3-pro | GPT-4.1 |
|---|---|---|
| Availability | Deprecated | Deprecated |
| Provider | OpenAI | OpenAI |
| Model type | Reasoning model | Chat model |
| First release | June 10, 2025 | April 14, 2025 |
| Context window | 200K tokens | 1M tokens |
| Max output | 100K tokens | 32K tokens |
| Architecture | Undisclosed | Undisclosed |
| Documented capabilities | Reasoning | Multilingual |
| Knowledge cutoff | 2024-06-01 | — |
| Input modalities | Text, Image | Text, Image |
| Output modalities | Text | Text |
| Weights | Closed / API only | Closed / API only |
| Commercial use | Proprietary | Proprietary |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used o3-pro or GPT-4.1? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.