Compare / head-to-head
Gemini 2.5 ProvsPhi-4 Multimodal Instruct
Gemini 2.5 Pro leads 1 of 1 shared benchmark. Gemini 2.5 Pro has the larger context window (1.0M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
1 – 0
Gemini 2.5 Pro leads
Cheaper per token
—
list price, input + output
Larger context
Gemini 2.5 Pro
1.0M tokens
Provider uptime (30d)
100% · —
Google · Microsoft
Gemini 2.5 Pro
Google · released 2025-06-17
textvisionaudio
Phi-4 Multimodal Instruct
Microsoft · released 2025-02-26
textvisionaudio
- Context
- 128K
- Max out
- 4K
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 19 · 0 core
Quality
Benchmark matrix
1 shared · 11 only Gemini 2.5 Pro · 18 only Phi-4 Multimodal Instruct| Benchmark | Gemini 2.5 Pro | Phi-4 Multimodal Instruct | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 1 | ||||
| MMMU | 82% ↗ | 55.1% ↗ | +26.9 pt | Gemini 2.5 Pro |
| Only Gemini 2.5 Pro reports · 11 | ||||
| SWE-bench Verified | 59.6% ↗ | not reported | — | — |
| GPQA Diamond | 86.4% ↗ | not reported | — | — |
| AIME 2025 | 88% ↗ | not reported | — | — |
| Humanity's Last Exam (no tools) | 21.6% ↗ | not reported | — | — |
| LMArena Elo | 1457.8 ↗ | not reported | — | — |
| Aider Polyglot | 82.2% ↗ | not reported | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | 0% ↗epoch | not reported | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 24.6% ↗epoch | not reported | — | — |
| LiveCodeBench | 74.2% ↗ | not reported | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 84.7% ↗epoch | not reported | — | — |
| SWE-bench Verified (Epoch AI run) | 57.6% ↗epoch | not reported | — | — |
| Only Phi-4 Multimodal Instruct reports · 18 | ||||
| AI2D | not reported | 82.3% ↗ | — | — |
| BLINK | not reported | 61.3% ↗ | — | — |
| ChartQA | not reported | 81.4% ↗ | — | — |
| DocVQA | not reported | 93.2% ↗ | — | — |
| InfoVQA | not reported | 72.7% ↗ | — | — |
| InterGPS | not reported | 48.6% ↗ | — | — |
| MathVista testmini | not reported | 62.4% ↗ | — | — |
| MMBench (dev-en) | not reported | 86.7% ↗ | — | — |
| MMMU-Pro (standard/vision) | not reported | 38.5% ↗ | — | — |
| OCRBench | not reported | 84.4% ↗ | — | — |
| POPE | not reported | 85.6% ↗ | — | — |
| ScienceQA Visual | not reported | 97.5% ↗ | — | — |
| Speech AI2D | not reported | 68.9% ↗ | — | — |
| Speech ChartQA | not reported | 69% ↗ | — | — |
| Speech DocVQA | not reported | 87.3% ↗ | — | — |
| Speech InfoVQA | not reported | 63.7% ↗ | — | — |
| TextVQA (val) | not reported | 75.6% ↗ | — | — |
| Video-MME (16 frames) | not reported | 55% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Gemini 2.5 Pro minus Phi-4 Multimodal Instruct in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Gemini 2.5 Pro | Phi-4 Multimodal Instruct | Edge |
|---|---|---|---|
| Context window | 1.0M tokens | 128K tokens | Gemini 2.5 Pro |
| Max output | 66K tokens | 4K tokens | Gemini 2.5 Pro |
| Input price / 1M | $1.25 ↗ | — | — |
| Output price / 1M | $10 ↗ | — | — |
| Cached input / 1M | $0.125 ↗ | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $11.25 | — | — |
| Modalities | text · vision · audio | text · vision · audio | Tie |
| Released | 2025-06-17 | 2025-02-26 | — |
| Cited benchmark scores | 14 | 19 | — |
Reliability
Provider status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
Gemini 2.5 Pro vs …
Models sharing the most benchmarksMore matchups
Phi-4 Multimodal Instruct vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Gemini 2.5 Pro and Phi-4 Multimodal Instruct on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.