Across 9 comparable benchmarks, Gemini 2.5-Pro has better reported results on 9 and worse on 0.
Moderate evidence9 of 9 shared results countedHow we compare
Gemini 2.5-Pro is no longer available through its official API. API prices for this version are not shown.
Price basis: Listed rates vs standard rates. The price section below shows the applicable periods.
| Fact | Gemini 2.5-Pro | GPT-4.1 mini |
|---|---|---|
| Context window | 1M tokens | 1M tokens |
| Input / 1M | — | $0.400 |
| Output / 1M | — | $1.60 |
| Blended (4:1) | — | $0.424 |
| Status | Retired | Deprecated |
| Released | June 5, 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
Gemini 2.5-Pro ahead on 3 of 3 counted
Gemini 2.5-Pro ahead on 1 of 1 counted
Gemini 2.5-Pro ahead on 1 of 1 counted
Gemini 2.5-Pro ahead on 1 of 1 counted
Gemini 2.5-Pro ahead on 1 of 1 counted
Gemini 2.5-Pro ahead on 2 of 2 counted
No fixed-scale average — see the rows in Benchmark results
No shared comparable results for Abstract reasoning, Computer & tool use, Research & search, Cybersecurity.
Published scores, with matching modes where available. Gaps show Gemini 2.5-Pro relative to GPT-4.1 mini. Reading the results
| Benchmark | Gemini 2.5-Pro | GPT-4.1 mini | Gap |
|---|---|---|---|
| Aider-PolyglotAgentic Development · Accuracy | 83.1Thinking · no tools · 32K | 32.4≠Normal · no tools | +50.7 pts#3 vs #32 of 47 |
| SWE-bench VerifiedRepository Engineering · Accuracy | 67.2Thinking · no tools | 23.6≠Normal · no tools | +43.6 pts#70 vs #106 of 108 |
| Terminal Bench HardAgentic Development · Accuracy | 26.5Thinking · tools | 7.6≠Normal · tools | +18.9 pts#76 vs #122 of 157 |
| FrontierMathMathematics · Accuracy | 11.0Normal · no tools | 4.5Normal · no tools | +6.5 pts#18 vs #30 of 48 |
| GPQA DiamondScientific Reasoning · Accuracy | 86.4Thinking · no tools | 65.0≠Normal · no tools | +21.4 pts#72 vs #157 of 200 |
| Fiction.liveBenchLong Reasoning · 16k-token score (%) | 91.7Normal · no tools | 44.4Normal · no tools | +47.3 pts#5 vs #25 of 28 |
| MMMU-ProVisual Understanding · Accuracy | 74.9Thinking · no tools | 58.7≠Normal · no tools | +16.2 pts#69 vs #108 of 136 |
| Creative WritingWriting | 1422Normal · no tools | 1145Normal · no tools | 1.24×#69 vs #93 of 121 |
| τ³-BankingService Workflows · Score | 13.7Thinking High · tools | 5.4≠Normal · tools | 2.53×#78 vs #107 of 120 |
Agentic Development · Accuracy
Repository Engineering · Accuracy
Agentic Development · Accuracy
Mathematics · Accuracy
Scientific Reasoning · Accuracy
Long Reasoning · 16k-token score (%)
Visual Understanding · Accuracy
Writing
Service Workflows · Score
A dash means no result is available.
| Benchmark | Gemini 2.5-Pro | GPT-4.1 mini |
|---|---|---|
| AIME 2024 | — | 49.6 |
| AIME2025 | 88.0 | — |
| CritPt | 2.6 | — |
| FrontierMath - Tier 4 | 2.1 | — |
| FrontierMath v2 | — | 6.7 |
| HLE | 21.6 | — |
| IMO-ProofBench | 55.2 | — |
| IMO-ProofBench Advanced | 17.6 | — |
| MATH-500 | 98.8 | — |
| MMLU | — | 87.5 |
| MMLU-Pro | 86.2 | — |
| SciCode | 46.3 | — |
| CodeClash | 1125 | — |
| GSO | 3.9 | — |
| IC SWE-Lancer (Diamond) | — | 33.0 |
| IOI 2024 (Vals v1) | 19 | — |
| IOI 2025 (Vals v1) | 15.2 | — |
| LiveCodeBench | 77.1 | — |
| SWE-Bench Pro - Commercial | 10.1 | — |
| Terminal Bench 2.0 | 32.6 | — |
| Terminal-Bench 1.0 | 25.3 | — |
| ECI | — | 135 |
| LiveBench | 58.3 | — |
| METR Time Horizons v1.1 | 38.7 | — |
| IF Bench | 49.0 | — |
| GDP.pdf | 10.2 | — |
| MMMU | 82.0 | — |
| VPCT | 46.4 | — |
| GDPval-AA | 22.0 | — |
| τ²-Bench | — | 53.0 |
| τ²-Bench - Telecom | — | 48.9 |
| ARC-AGI-1 | 37.0 | — |
| ARC-AGI-2 | 4.9 | — |
| SimpleBench | 62.4 | — |
| BrowseComp | 7.8 | — |
| SimpleQA | 54.0 | — |
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.
Gemini 2.5-Pro: Listed text rates; pricing tier not confirmed.
GPT-4.1 mini: Standard pricing. OpenAI official pricing (developers.openai.com/api/docs/pricing) verified 2026-11-13; standard tier; Deprecated; shutdown 2027-01-14 (replacement: gpt-6-luna)
| Price / 1M tokens | Gemini 2.5-Pro | GPT-4.1 mini | Ratio |
|---|---|---|---|
| Input | — | $0.40 | — |
| Output | — | $1.60 | — |
| Cached input | — | $0.10 | — |
| Cache write | — | — | — |
| Blended | — | $0.424 | — |
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.
| Benchmark | Gemini 2.5-Pro | GPT-4.1 mini |
|---|---|---|
| ARC-AGI-2 | Not reported |
| Attribute | Gemini 2.5-Pro | GPT-4.1 mini |
|---|---|---|
| Availability | Retired | Deprecated |
| Provider | Google Deep Mind | OpenAI |
| Model type | Reasoning model | Chat model |
| First release | June 5, 2025 | April 14, 2025 |
| Context window | 1M tokens | 1M tokens |
| Max output | 64K tokens | 32K tokens |
| Architecture | Undisclosed | Undisclosed |
| Documented capabilities | Reasoning, Multilingual | Multilingual |
| 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 Gemini 2.5-Pro or GPT-4.1 mini? 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.