Compare models
Pick up to 4 models for a side-by-side spec sheet. Hover any dotted metric name for what the test measures. Best value in each row is highlighted.
Gemini 3.6 FlashGPT-5.4
| Gemini 3.6 Flash | GPT-5.4 OpenAI | |
|---|---|---|
| Pricing | ||
| Input $/1M | $1.50 | $2.50 |
| Output $/1M | $7.50 | $15.00 |
| Blended $/1M (3:1) | $3.00 | $5.63 |
| Cache read $/1M | $0.15 | $0.25 |
| Cache write $/1M | $0.083 | — |
| Quality indices | ||
| Intelligence Index | 50.1 | 51.4 |
| Coding Index | 69.2 | 71.1 |
| Math Index | — | — |
| Agentic Index | 38.7 | 41.1 |
| Speed & latency | ||
| Output speed (tok/s) | 301 | 152 |
| Time to first token | 11.80s | 129.43s |
| Time to first answer token | 11.80s | 129.43s |
| Benchmarks (Artificial Analysis) | ||
| GPQA Diamond | 92.8% | 92.0% |
| Humanity's Last Exam | 38.3% | 41.6% |
| SciCode | 52.7% | 56.6% |
| IFBench | — | 73.9% |
| AA-LCR (Long Context Reasoning) | 69.7% | 74.0% |
| Terminal-Bench Hard | — | 57.6% |
| Terminal-Bench 2.1 | 77.5% | 78.3% |
| τ²-Bench (Telecom) | — | 87.1% |
| τ³-Bench Banking | 24.5% | 30.3% |
| Design Arena (Elo) | ||
| Website | 1,332 | 1,247 |
| Web apps | — | 1,122 |
| Full-stack | — | 1,076 |
| Mobile apps | — | 1,147 |
| Android native | — | 1,020 |
| UI components | — | 1,281 |
| Data viz | — | 1,268 |
| SVG | — | 1,241 |
| 3D | — | 1,160 |
| Game dev | — | 1,298 |
| Godot game dev | — | 1,135 |
| Code categories | — | 1,245 |
| ASCII art | — | 1,241 |
| Specs | ||
| Context window | 1.05M | 1.05M |
| Max output tokens | 66K | 128K |
| Input modalities | text, image, video, file, audio | text, image, file |
| Tokenizer | Gemini | GPT |