Compare / Gemini 3.8 Flash vs GPT-5.6 Terra
Gemini 3.8 Flash vs GPT-5.6 Terra
27 results from 7 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Gemini 3.8 Flash | GPT-5.6 Terra |
|---|---|---|---|
| APEX-Agentsreported | Tasks passed% of tasks | 64.3% | 58.2% |
| Arena (formerly LMArena)reported | Confirmed task successIPS effect estimate | 0.09 | -0.06 |
| Arena (formerly LMArena)reported | Praise over complaintIPS effect estimate | 0.05 | 0.03 |
| Arena (formerly LMArena)reported | SteerabilityIPS effect estimate | 0.04 | 0.04 |
| Arena (formerly LMArena)reported | Tool groundingIPS effect estimate | 0.00 | 0.00 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,477 | 1,467 |
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,482 | 1,466 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,482 | 1,428 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,529 | 1,504 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,488 | 1,462 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,492 | 1,465 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,486 | 1,445 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 76.3 | 76.7 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 40.9 | 42.1 |
| Artificial Analysisreported | GPQA Diamond% of questions | 95.3% | 92.5% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 47.8% | 42.9% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 84.0% | 83.0% |
| Artificial Analysisreported | SciCode% of problems | 56.6% | 55.0% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 87.6% | 88.0% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 19.7% | 35.4% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 45.8% | 40.2% |
| GDP.pdfreported | Professional document tasks% of rubric | 23.4% | 24.7% |
| SimpleQA Verified (Epoch AI)reported | Correct answers% of questions | 69.7% | 43.2% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 13.0% | 22.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.34 | 0.39 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 78.5 | 60.4 |
| Vending-Bench 2reported | Money after a yearUS dollars | $5,093.79 | $7,343.21 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.