Model comparison
Gemini 3.8 Flash vs GPT-5.4 Pro
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 58.9 on the Noometry Index.
Last verified . 11 shared benchmarks.
GoogleSummary
- They share 11 benchmarks with published results for both. Gemini 3.8 Flash scores higher in 3 categories and GPT-5.4 Pro in 1 category; 4 gaps are clear of the uncertainty.
- The widest gap is in coding, where Gemini 3.8 Flash leads 59.2 to 43.2.
- The biggest single-benchmark swing is FrontierMath Tier 4: 22% for Gemini 3.8 Flash and 58.5% for GPT-5.4 Pro.
- Gemini 3.8 Flash is cheaper at $0.75 / $3.75 per million input/output tokens, against $30 / $180 for GPT-5.4 Pro.
- GPT-5.4 Pro accepts more context: 1.05M tokens versus 1.05M.
Side by side
| Gemini 3.8 Flash | GPT-5.4 Pro | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 61.8 | 58.9 |
| Released | 2026-09-02 | 2026-03-05 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 66K | 128K |
| Input $ / M tokens | $0.75 | $30 |
| Output $ / M tokens | $3.75 | $180 |
| Results tracked | 50 | 16 |
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-5.4 Pro: 43.2 (#90)
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| WeirdML | 84.8% | 57.4% |
| DeepSWE | 73.8% | — |
| FrontierCode | 41.2% | — |
| CursorBench | 39.6% | — |
| LMArena WebDev | 1584 | — |
| FrontierSWE | 19.6% | — |
| SciCode | 56.6% | — |
| LMArena Coding | 1510 | — |
| ALE-Bench | 1,270 | — |
Agentic & Tool Use Not comparable
Gemini 3.8 Flash: 41.8 (#21), GPT-5.4 Pro: —
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| APEX-Agents | 64.3% | — |
| Remote Labor Index | 5.8% | — |
| GDP.pdf | 23.4% | — |
| Vending-Bench 2 | 5,094 | — |
Reasoning Gemini 3.8 Flash leads
Gemini 3.8 Flash: 76.9 (#5), GPT-5.4 Pro: 70.7 (#13)
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| ARC-AGI-2 | 89.2% | 83.3% |
| ARC-AGI-1 | 98.5% | 94.5% |
| CritPt | 18.3% | 30% |
| Chess Puzzles | 61% | 58.6% |
| Epoch Capabilities Index | 156.71 | 158.93 |
| SimpleBench | — | 74.1% |
| NYT Connections (extended) | 97.4% | — |
| EnigmaEval | — | 23.8% |
| LMArena Hard Prompts | 1508 | — |
| Mystery Game Puzzles | 47% | — |
| DTBench | 95.7% | — |
| LMCA | 52.9% | — |
| Surface Evolver Bench | 76.9% | — |
Math GPT-5.4 Pro leads
Gemini 3.8 Flash: 65.3 (#28), GPT-5.4 Pro: 72.4 (#22)
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| FrontierMath (Tiers 1-3) | 68.4% | 82.5% |
| FrontierMath Tier 4 | 22% | 58.5% |
| OTIS Mock AIME 2024-2025 | 98.9% | — |
| ProofBench | 48% | — |
| LMArena Math | 1528 | — |
| FrontierMath (Feb 2025 set) | — | 50% |
| FrontierMath Tier 4 (v1) | — | 37.5% |
Knowledge Gemini 3.8 Flash leads
Gemini 3.8 Flash: 74.8 (#2), GPT-5.4 Pro: 68.3 (#7)
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| GPQA Diamond | 95.4% | 94.6% |
| Humanity's Last Exam | 44.5% | 44.3% |
| SimpleQA Verified | 69.7% | 46.3% |
| Vectara Hallucination Rate | — | 8.3% |
| LMArena Expert | 1524 | — |
Multimodal Not comparable
Gemini 3.8 Flash: 40.7 (#45), GPT-5.4 Pro: —
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| LMArena Vision | 1314 | — |
| Blueprint-Bench 2 | 38.6% | — |
| Furniture Assembly | 31.7% | — |
Multilingual Not comparable
Gemini 3.8 Flash: 58.0 (#5), GPT-5.4 Pro: —
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| LMArena Non-English | 1491 | — |
| LMArena Chinese | 1554 | — |
| LMArena French | 1498 | — |
| LMArena German | 1493 | — |
| LMArena Japanese | 1502 | — |
| LMArena Korean | 1459 | — |
| LMArena Russian | 1515 | — |
| LMArena Spanish | 1485 | — |
Instruction Following Not comparable
Gemini 3.8 Flash: 78.0 (#13), GPT-5.4 Pro: —
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| LMArena Instruction Following | 1490 | — |
Long Context Not comparable
Gemini 3.8 Flash: 46.3 (#24), GPT-5.4 Pro: —
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| LMArena Longer Query | 1508 | — |
Writing & Preference Not comparable
Gemini 3.8 Flash: 72.2 (#15), GPT-5.4 Pro: —
| Benchmark | Gemini 3.8 Flash | GPT-5.4 Pro |
|---|---|---|
| LMArena Text | 1499 | — |
| LMArena Creative Writing | 1492 | — |
| EQ-Bench Creative Writing | 1748 | — |
| LMArena Multi-Turn | 1501 | — |
Frequently asked questions
Is Gemini 3.8 Flash better than GPT-5.4 Pro?
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 58.9 on the Noometry Index.
Which is cheaper, Gemini 3.8 Flash or GPT-5.4 Pro?
Gemini 3.8 Flash is cheaper. It lists at $0.75 per million input tokens and $3.75 per million output tokens; GPT-5.4 Pro lists at $30 and $180.
Is Gemini 3.8 Flash or GPT-5.4 Pro better for coding?
Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 43.2 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 Pro does, with 1.05M tokens against 1.05M.
How many benchmarks do Gemini 3.8 Flash and GPT-5.4 Pro share?
11 benchmarks have published results for both models. Gemini 3.8 Flash has 50 scored results on Noometry and GPT-5.4 Pro has 16.