Across 19 comparable benchmarks, Gemini 3.1 Pro Preview has better reported results on 19 and worse on 0.
Moderate evidence19 of 19 shared results countedHow we compare
Price basis: Listed rates vs standard rates. The price section below shows the applicable periods.
| Fact | Gemini 3.1 Pro Preview | GPT OSS 120B |
|---|---|---|
| Context window | 1M tokens | 128K tokens |
| Input / 1M | $2.00 | $0.350 |
| Output / 1M | $12.00 | $0.750 |
| Blended (4:1) | $2.70 | $0.430 |
| Status | Preview | Available |
| Released | February 20, 2026 | August 6, 2025 |
Blended is 4:1 input to output with 90% cache hits at listed rates β set your own mix.
Mean of 9 shared benchmarks grouped by capability on a 0β100 scale. How averages are built
Gemini 3.1 Pro Preview ahead on 4 of 4 counted
Gemini 3.1 Pro Preview ahead on 1 of 1 counted
No fixed-scale average β see the rows in Benchmark results
Gemini 3.1 Pro Preview ahead on 3 of 3 counted
Gemini 3.1 Pro Preview ahead on 3 of 3 counted
Gemini 3.1 Pro Preview ahead on 1 of 1 counted
Gemini 3.1 Pro Preview ahead on 4 of 4 counted
No shared comparable results for Mathematics, Research & search, Long context, Cybersecurity.
Published scores, with matching modes where available. Gaps show Gemini 3.1 Pro Preview relative to GPT OSS 120B. Reading the results
| Benchmark | Gemini 3.1 Pro Preview | GPT OSS 120B | Gap |
|---|---|---|---|
| APEX-SWERepository Engineering Β· Pass@1 | 33.9Thinking High Β· tools | 3.3Thinking High Β· tools | +30.6 pts#45 vs #61 of 61 |
| LiveCodeBenchAlgorithmic Coding Β· Pass @K | 91.7Thinking High Β· tools | 87.8β Thinking High Β· no tools | 1.04Γ#4 vs #14 of 113 |
| SWE-bench VerifiedRepository Engineering Β· Accuracy | 80.6Thinking High Β· tools | 60.1β Thinking Β· no tools | +20.5 pts#12 vs #78 of 108 |
| Terminal Bench HardAgentic Development Β· Accuracy | 53.8Thinking Β· tools | 23.5β Thinking High Β· tools | +30.3 pts#8 vs #88 of 157 |
| SimpleBenchCommonsense Β· Score (AVG@5) | 79.6Normal Β· no tools | 22.1β Thinking Β· no tools | 3.60Γ#15 vs #99 of 106 |
| CritPtScientific Reasoning Β· Score | 17.7Thinking Β· no tools | 1.1β Thinking High Β· no tools | 16.1Γ#34 vs #103 of 134 |
| GPQA DiamondScientific Reasoning Β· Accuracy | 94.4Reported best | 80.1β Thinking Β· no tools | +14.3 pts#12 vs #111 of 200 |
| SciCodeScientific Computing Β· Score | 58.7Thinking Β· no tools | 34.0β Thinking High Β· no tools | 1.73Γ#18 vs #88 of 98 |
| APEX-AgentsTool Orchestration Β· Score | 35.1Reported best | 4.4Reported best | 7.98Γ#52 vs #61 of 61 |
| Pinch BenchTool Orchestration Β· Accuracy | 86.7Thinking Β· tools | 60.6Thinking Β· tools | +26.1 pts#11 vs #36 of 38 |
| PinchBench v2Tool Orchestration Β· Average score (%) | 81.0Reported best | 44.8Reported best | +36.2 pts#13 vs #46 of 51 |
| GDP.pdfDocuments & Charts Β· Task pass rate (%) | 17.8Thinking Β· no tools | 4.0β Thinking High Β· no tools | +13.8 pts#33 vs #77 of 96 |
| Creative WritingWriting | 1491Normal Β· no tools | 956Normal Β· no tools | 1.56Γ#61 vs #104 of 121 |
| AutomationBenchOffice & Business Β· Pass Rate | 8.7Thinking High Β· toolsΒ· zapier-private-heldout | 0.15Thinking High Β· toolsΒ· zapier-private-heldout | +8.5 pts#51 vs #64 of 64 |
| Harvey Lab-AALegal Β· Criterion pass rate (some entries report all-pass rate) | 58.9Thinking Β· tools | 13.9β Thinking High Β· tools | +45.0 pts#33 vs #39 of 39 |
| ΟΒ³-BankingService Workflows Β· Score | 26.0Thinking High Β· tools | 12.8Thinking High Β· tools | 2.03Γ#46 vs #83 of 120 |
| HLEKnowledge Exams Β· Accuracy | 51.4Thinking High Β· tools | 19.0β Thinking Β· tools | +32.4 pts#44 vs #107 of 151 |
| ECICapability Indices Β· ECI score (capability index, higher is better) | 155Reported best | 140Reported best | 1.11Γ#31 vs #97 of 168 |
| LiveBenchCross-capability Suites Β· Accuracy | 77.0Thinking High Β· no tools | 46.1β Normal Β· no tools | +30.9 pts#24 vs #100 of 106 |
Repository Engineering Β· Pass@1
Algorithmic Coding Β· Pass @K
Repository Engineering Β· Accuracy
Agentic Development Β· Accuracy
Commonsense Β· Score (AVG@5)
Scientific Reasoning Β· Score
Scientific Reasoning Β· Accuracy
Scientific Computing Β· Score
Tool Orchestration Β· Score
Tool Orchestration Β· Accuracy
Tool Orchestration Β· Average score (%)
Documents & Charts Β· Task pass rate (%)
Writing
Office & Business Β· Pass Rate
Legal Β· Criterion pass rate (some entries report all-pass rate)
Service Workflows Β· Score
A dash means no result is available.
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 Β· 5% steps
Blended price = effective input rate at 90% cache hits Γ 80% + output rate Γ 20%
Gemini 3.1 Pro Preview costs529%morethan GPT OSS 120B per million tokens β 6.3Γ the price.Using listed rates vs standard rates at this token mix.
Gemini 3.1 Pro Preview: Standard pricing. context_length <= 200K Google Gemini API official pricing (ai.google.dev/gemini-api/docs/pricing) verified 2026-10-21; paid tier; standard tier; prompts <= 200k tokens context_length > 200K Google Gemini API official pricing (ai.google.dev/gemini-api/docs/pricing) verified 2026-10-21; paid tier; standard tier; prompts > 200k tokens
GPT OSS 120B: Standard pricing. Cerebras Shared Inference official pricing (gpt-oss-120b) verified 2026-11-13
| Price / 1M tokens | Gemini 3.1 Pro Preview | GPT OSS 120B | Ratio |
|---|---|---|---|
| Input | $2.00 | $0.35 | 5.71Γ |
| Output | $12.00 | $0.75 | 16.0Γ |
| Cached input | $0.20 | β | β |
| Cache write | β | β | β |
| Blended | $2.70 | $0.43 | 6.29Γ |
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.
Gemini 3.1 Pro Preview relative to GPT OSS 120B
Summary: 1 of 1 shared benchmarks have matching recorded settings and sources. The average is a geometric mean of 1 positive-cost ratios, one equal weight per benchmark. This benchmark set can include related tests and indexes; it is not a workload cost estimate.
Gemini 3.1 Pro Preview 30Γ more expensive
Gemini 3.1 Pro Preview 30Γ more expensive
Effort, protocol, aggregation and source match; other run conditions may differ.
Gemini 3.1 Pro Preview
GPT OSS 120B
| Benchmark | Gemini 3.1 Pro Preview | GPT OSS 120B |
|---|---|---|
| AA Intelligence Index v5.0.1 | Not reported | |
| Agents' Last Exam v1.0 | $14.2116 / task Average Β· Pass rate: 16.4 Unspecified Reasoning (With Tools) Β· pass-rate | Not reported |
| AIME 2026 | Not reported | |
| DeepSWE | Not reported | |
| IOI (Vals v2) | Not reported | |
| LiveBench 2.0 | $0.6982 / task Average Β· Overall score (equal-weight mean of 7 suite scores, 0-100): 54.9 Thinking High (With Tools) | Not reported |
| MysteryMechanism | Not reported | |
| Tax Agent Bench | $0.3016 / task Average Β· Accuracy (%, rubric Γ citation quality): 9.3 Thinking High (With Tools) | Not reported |
| Vals Index v2.1 | Not reported | |
| Vibe Code Bench 1-100 | Not reported |
| Attribute | Gemini 3.1 Pro Preview | GPT OSS 120B |
|---|---|---|
| Availability | Preview | Available |
| Provider | Google Deep Mind | OpenAI |
| Model type | Multimodal model | Reasoning model |
| First release | February 20, 2026 | August 6, 2025 |
| Context window | 1M tokens | 128K tokens |
| Max output | 64K tokens | 128K tokens |
| Total parameters | β | 117B |
| Active parameters | β | 5.1B |
| Architecture | Undisclosed | Mixture of Experts (MoE) |
| Documented capabilities | Reasoning, Multilingual | Reasoning, Multilingual |
| Knowledge cutoff | 2025-01 | β |
| Input modalities | Text, Image, Audio, Video | Text |
| Output modalities | Text | Text |
| Weights | Closed / API only | Open weights |
| Licence | β | Apache 2.0 |
| Commercial use | Proprietary | Commercial use allowed |
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 3.1 Pro Preview or GPT OSS 120B? 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.