Across 14 comparable benchmarks, OpenAI o3 has better reported results on 9 and worse on 5.
Moderate evidence14 of 14 shared results countedHow we compare
Gemini 2.5 Pro Experimental 03-25 is no longer available through its official API. API prices for this version are not shown.
Price basis: Standard rates vs listed rates. The price section below shows the applicable periods.
| Fact | OpenAI o3 | Gemini 2.5 Pro Experimental 03-25 |
|---|---|---|
| Context window | 200K tokens | 1M tokens |
| Input / 1M | $2.00 | — |
| Output / 1M | $8.00 | — |
| Blended (4:1) | $2.12 | — |
| Status | Deprecated | Retired |
| Released | April 16, 2025 | March 25, 2025 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 9 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
OpenAI o3 ahead on 3 of 3 counted
OpenAI o3 ahead on 1 of 1 counted
No fixed-scale average — see the rows in Benchmark results
Gemini 2.5 Pro Experimental 03-25 ahead on 2 of 3 counted
Gemini 2.5 Pro Experimental 03-25 ahead on 1 of 1 counted
OpenAI o3 ahead on 1 of 1 counted
Split 1–1 of 2 counted
OpenAI o3 ahead on 1 of 1 counted
No fixed-scale average — see the rows in Benchmark results
No shared comparable results for Computer & tool use, Research & search, Cybersecurity.
Published scores, with matching modes where available. Gaps show OpenAI o3 relative to Gemini 2.5 Pro Experimental 03-25. Reading the results
| Benchmark | OpenAI o3 | Gemini 2.5 Pro Experimental 03-25 | Gap |
|---|---|---|---|
| Aider-PolyglotAgentic Development · Accuracy | 81.3Thinking High · no tools | 72.9≠Normal · no tools | +8.4 pts#4 vs #8 of 47 |
| LiveCodeBenchAlgorithmic Coding · Pass @K | 75.8Normal · no tools | 70.4Normal · no tools | 1.08×#43 vs #55 of 113 |
| SWE-bench VerifiedRepository Engineering · Accuracy | 69.1Thinking · no tools | 63.8≠Normal · no tools | +5.3 pts#65 vs #75 of 108 |
| SimpleBenchCommonsense · Score (AVG@5) | 53.1Thinking High · no tools | 51.6≠Normal · no tools | 1.03×#61 vs #63 of 106 |
| AIME 2024Mathematics · Accuracy | 91.6Normal · no tools | 92.0Normal · no tools | −0.4 pts#9 vs #8 of 54 |
| AIME2025Mathematics · Accuracy | 88.9Thinking · no tools | 86.9≠Normal · no tools | +2.0 pts#37 vs #43 of 86 |
| FrontierMath - Tier 4Mathematics · Accuracy | 2.1Thinking High · no tools | 4.2≠Normal · no tools | −2.1 pts#31 vs #21 of 50 |
| GPQA DiamondScientific Reasoning · Accuracy | 83.3Thinking · no tools | 84.0≠Normal · no tools | −0.7 pts#94 vs #88 of 200 |
| Fiction.liveBenchLong Reasoning · 16k-token score (%) | 88.9Thinking Medium · no tools | 66.7≠Normal · no tools | +22.2 pts#6 vs #15 of 28 |
| GeoBench ACWVisual Understanding · ACW country accuracy (%) | 74.0Thinking Medium · no tools | 81.0≠Normal · tools + web | −7.0 pts#16 vs #10 of 32 |
| VPCTVisual Understanding · Accuracy (%) | 52.0Thinking Medium · no tools | 48.0≠Normal · no tools | +4.0 pts#7 vs #8 of 26 |
| Creative WritingWriting | 1675Normal · no tools | 1396Normal · no tools | 1.20×#37 vs #76 of 121 |
| Reported separately from the capability domains; not included in the averages. Read each metric’s direction — lower can be better. | |||
| SimpleQAHonesty & Factuality · Accuracy | 49.4Normal · no tools | 52.9Normal · no tools | −3.5 pts#14 vs #12 of 46 |
| HLEKnowledge Exams · Accuracy | 20.3Thinking High · no tools | 18.8≠Normal · no tools | +1.5 pts#102 vs #108 of 151 |
Agentic Development · Accuracy
Algorithmic Coding · Pass @K
Repository Engineering · Accuracy
Commonsense · Score (AVG@5)
Mathematics · Accuracy
Mathematics · Accuracy
Mathematics · Accuracy
Scientific Reasoning · Accuracy
Long Reasoning · 16k-token score (%)
Visual Understanding · ACW country accuracy (%)
Visual Understanding · Accuracy (%)
Writing
Honesty & Factuality · Accuracy
Knowledge Exams · Accuracy
A dash means no result is available.
| Benchmark | OpenAI o3 | Gemini 2.5 Pro Experimental 03-25 |
|---|---|---|
| CritPt | 1.1 | — |
| FrontierMath | 10.3 | — |
| FrontierMath v2 | 33.3 | — |
| IMO-ProofBench | 20.5 | — |
| IMO-ProofBench Advanced | 20.5 | — |
| MATH-500 | 99.2 | — |
| MMLU-Pro | 85.6 | — |
| CodeClash | 1343 | — |
| GSO | 8.8 | — |
| Terminal Bench Hard | 37.1 | — |
| Terminal-Bench 1.0 | 30.2 | — |
| WeirdML v2 | 52.4 | — |
| Claw Bench | — | 80.4 |
| OSWorld-Verified | 23.0 | — |
| Pinch Bench | — | 71.9 |
| ECI | 147 | — |
| METR Time Horizons v1.1 | 91.3 | — |
| MMMU | 82.9 | — |
| MMMU-Pro | 76.4 | — |
| SAGE | 41.8 | — |
| ARC-AGI-1 | 60.8 | — |
| ARC-AGI-2 | 6.5 | — |
| BALROG | — | 43.3 |
| GAIA | 28.5 | — |
| MedCode | 47.3 | — |
| MedScribe | 76.7 | — |
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%
A blended cost needs a published input and output price on both sides, and one of these two does not have them yet.
OpenAI o3: 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)
Gemini 2.5 Pro Experimental 03-25: Listed text rates; pricing tier not confirmed.
| Price / 1M tokens | OpenAI o3 | Gemini 2.5 Pro Experimental 03-25 | Ratio |
|---|---|---|---|
| Input | $2.00 | — | — |
| Output | $8.00 | — | — |
| Cached input | $0.50 | — | — |
| Cache write | — | — | — |
| Blended | $2.12 | — | — |
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 | OpenAI o3 | Gemini 2.5 Pro Experimental 03-25 |
|---|---|---|
| Availability | Deprecated | Retired |
| Provider | OpenAI | Google Deep Mind |
| Model type | Reasoning model | Reasoning model |
| First release | April 16, 2025 | March 25, 2025 |
| Context window | 200K tokens | 1M tokens |
| Max output | 100K tokens | 64K tokens |
| Architecture | Undisclosed | Undisclosed |
| Documented capabilities | Reasoning, Multilingual | Reasoning, Multilingual |
| Input modalities | Text, Image | Text, Image, Audio, Video |
| 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 OpenAI o3 or Gemini 2.5 Pro Experimental 03-25? 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.