Model comparison
Gemini 3.8 Flash vs GPT-5.5
GPT-5.5 is the stronger model overall, scoring 63.4 to 61.8 on the Noometry Index. Gemini 3.8 Flash costs 7.5× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.
Last verified . 47 shared benchmarks.
GoogleSummary
- They share 47 benchmarks with published results for both. Gemini 3.8 Flash scores higher in 5 categories and GPT-5.5 in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.5 leads 81.7 to 65.3.
- The biggest single-benchmark swing is FrontierMath Tier 4: 22% for Gemini 3.8 Flash and 72.5% for GPT-5.5.
- Gemini 3.8 Flash is cheaper at $0.75 / $3.75 per million input/output tokens, against $5 / $30 for GPT-5.5.
- GPT-5.5 accepts more context: 1.05M tokens versus 1.05M.
Side by side
| Gemini 3.8 Flash | GPT-5.5 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 61.8 | 63.4 |
| Released | 2026-09-02 | 2026-04-23 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 66K | 128K |
| Input $ / M tokens | $0.75 | $5 |
| Output $ / M tokens | $3.75 | $30 |
| Results tracked | 50 | 71 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3.8 Flash leads
Gemini 3.8 Flash: 59.2 (#15), GPT-5.5: 58.2 (#17)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| DeepSWE | 73.8% | 67% |
| FrontierCode | 41.2% | 43% |
| LMArena WebDev | 1584 | 1513 |
| SciCode | 56.6% | 56.1% |
| WeirdML | 84.8% | 84.9% |
| LMArena Coding | 1510 | 1494 |
| ALE-Bench | 1,270 | 1,943 |
| SWE-bench Verified | — | 80.6% |
| CursorBench | 39.6% | — |
| FrontierSWE | 19.6% | — |
| GSO | — | 40.2% |
| MirrorCode | — | 10% |
Agentic & Tool Use GPT-5.5 leads
Gemini 3.8 Flash: 41.8 (#21), GPT-5.5: 50.7 (#6)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| APEX-Agents | 64.3% | 55.1% |
| Remote Labor Index | 5.8% | 6.3% |
| GDP.pdf | 23.4% | 26% |
| Vending-Bench 2 | 5,094 | 7,524 |
| Terminal-Bench | — | 84.7% |
| OSWorld 2.0 | — | 13% |
| τ²-bench Banking | — | 44.6% |
| DeepResearch Bench | — | 54% |
| PostTrainBench | — | 27.2% |
| ExploitBench | — | 47.4% |
| GBAEval | — | 53.2% |
| LMArena Search | — | 1242 |
Reasoning Gemini 3.8 Flash leads
Gemini 3.8 Flash: 76.9 (#5), GPT-5.5: 72.8 (#11)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| ARC-AGI-2 | 89.2% | 85% |
| NYT Connections (extended) | 97.4% | 96.2% |
| ARC-AGI-1 | 98.5% | 95% |
| CritPt | 18.3% | 27.1% |
| Chess Puzzles | 61% | 54% |
| LMArena Hard Prompts | 1508 | 1489 |
| Mystery Game Puzzles | 47% | 56% |
| DTBench | 95.7% | 96% |
| LMCA | 52.9% | 54.3% |
| Surface Evolver Bench | 76.9% | 88.1% |
| Epoch Capabilities Index | 156.71 | 159.1 |
| SimpleBench | — | 69% |
| Kagi LLM Benchmark | — | 88.8% |
| EBR-Bench | — | 34.3% |
| Bench to the Future 3 | — | 0.14 |
| ForecastBench | — | 60.6 |
Math GPT-5.5 leads
Gemini 3.8 Flash: 65.3 (#28), GPT-5.5: 81.7 (#11)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 68.4% | 85.3% |
| FrontierMath Tier 4 | 22% | 72.5% |
| OTIS Mock AIME 2024-2025 | 98.9% | 100% |
| ProofBench | 48% | 50% |
| LMArena Math | 1528 | 1486 |
| MathArena Final-Answer Competitions | — | 94.3% |
| FrontierMath (Feb 2025 set) | — | 51.7% |
| FrontierMath Erdős | — | 0% |
| FrontierMath Tier 4 (v1) | — | 35.4% |
Knowledge Gemini 3.8 Flash leads
Gemini 3.8 Flash: 74.8 (#2), GPT-5.5: 64.4 (#17)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| GPQA Diamond | 95.4% | 94% |
| SimpleQA Verified | 69.7% | 63% |
| LMArena Expert | 1524 | 1508 |
| Humanity's Last Exam | 44.5% | — |
| Vectara Hallucination Rate | — | 9.3% |
Multimodal GPT-5.5 leads
Gemini 3.8 Flash: 40.7 (#45), GPT-5.5: 46.9 (#12)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| LMArena Vision | 1314 | 1297 |
| Blueprint-Bench 2 | 38.6% | 36.2% |
| Furniture Assembly | 31.7% | 44.2% |
| LMArena Document | — | 1486 |
Multilingual Gemini 3.8 Flash leads
Gemini 3.8 Flash: 58.0 (#5), GPT-5.5: 56.4 (#20)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| LMArena Non-English | 1491 | 1467 |
| LMArena Chinese | 1554 | 1533 |
| LMArena French | 1498 | 1486 |
| LMArena German | 1493 | 1480 |
| LMArena Japanese | 1502 | 1498 |
| LMArena Korean | 1459 | 1460 |
| LMArena Russian | 1515 | 1473 |
| LMArena Spanish | 1485 | 1468 |
Instruction Following Too close to call
Gemini 3.8 Flash: 78.0 (#13), GPT-5.5: 77.5 (#18)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| LMArena Instruction Following | 1490 | 1479 |
Long Context GPT-5.5 leads
Gemini 3.8 Flash: 46.3 (#24), GPT-5.5: 48.3 (#12)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| LMArena Longer Query | 1508 | 1484 |
| CL-bench Life | — | 22.2% |
Writing & Preference Too close to call
Gemini 3.8 Flash: 72.2 (#15), GPT-5.5: 72.7 (#13)
| Benchmark | Gemini 3.8 Flash | GPT-5.5 |
|---|---|---|
| LMArena Text | 1499 | 1472 |
| LMArena Creative Writing | 1492 | 1455 |
| EQ-Bench Creative Writing | 1748 | 1844 |
| LMArena Multi-Turn | 1501 | 1476 |
| EQ-Bench 4 | — | 1315 |
Frequently asked questions
Is Gemini 3.8 Flash better than GPT-5.5?
GPT-5.5 is the stronger model overall, scoring 63.4 to 61.8 on the Noometry Index. Gemini 3.8 Flash costs 7.5× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.
Which is cheaper, Gemini 3.8 Flash or GPT-5.5?
Gemini 3.8 Flash is cheaper. It lists at $0.75 per million input tokens and $3.75 per million output tokens; GPT-5.5 lists at $5 and $30.
Is Gemini 3.8 Flash or GPT-5.5 better for coding?
Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 58.2 in the Noometry coding category.
Which has the bigger context window?
GPT-5.5 does, with 1.05M tokens against 1.05M.
How many benchmarks do Gemini 3.8 Flash and GPT-5.5 share?
47 benchmarks have published results for both models. Gemini 3.8 Flash has 50 scored results on Noometry and GPT-5.5 has 71.