Gemini 4 Pro (Projected) vs GPT-Image-2.5 (Sunburst)
A direct, empirical evaluation of Gemini 4 Pro (Projected) (Google DeepMind) against GPT-Image-2.5 (Sunburst) (OpenAI).
Gemini 4 Pro (Projected)
Developed by Google DeepMind · Proprietary Omni-Modal Unified Diffusion Transformer parameters
Repository-scale 10M-token refactoring, automated formal theorem proving, parametric SVG and 3D GLTF asset synthesis, and zero-hallucination execution.
GPT-Image-2.5 (Sunburst)
Developed by OpenAI · Proprietary Multimodal DiT parameters
High-fidelity typographic rendering, raytracing reflection accuracy, sketch-to-image conditioning, and localized comment inpainting.
| Metric | Gemini 4 Pro (Projected) | GPT-Image-2.5 (Sunburst) |
|---|---|---|
| SWE-bench Verified | 75.8% | 62.4% |
| MMLU-Pro | 92.4% | 89.5% |
| MATH-500 | 97.2% | 91.2% |
| Context Window | 2,000,000 - 10,000,000 tokens | Spatial Vector Token Canvas (1792×1024) |
| Input Pricing (per 1M) | $1.50 | $2.50 |
| Output Pricing (per 1M) | $7.50 | $10.00 |