Model comparison
Gemini 3.8 Flash vs Phi-4
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 31.2 on the Noometry Index. Phi-4 costs 17× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.
Last verified . 21 shared benchmarks.
Google
MicrosoftSummary
- They share 21 benchmarks with published results for both. Gemini 3.8 Flash scores higher in 9 categories and Phi-4 in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.8 Flash leads 76.9 to 17.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 98.9% for Gemini 3.8 Flash and 13.8% for Phi-4.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.75 / $3.75 for Gemini 3.8 Flash.
- Gemini 3.8 Flash accepts more context: 1.05M tokens versus 128K.
- Phi-4 has downloadable open weights; the other is API-only.
Side by side
| Gemini 3.8 Flash | Phi-4 | |
|---|---|---|
| Provider | Microsoft | |
| Noometry Index | 61.8 | 31.2 |
| Released | 2026-09-02 | 2024-12-11 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 128K |
| Max output | 66K | 4K |
| Input $ / M tokens | $0.75 | $0.07 |
| Output $ / M tokens | $3.75 | $0.14 |
| Results tracked | 50 | 37 |
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), Phi-4: 34.4 (#239)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| LMArena Coding | 1510 | 1231 |
| DeepSWE | 73.8% | — |
| FrontierCode | 41.2% | — |
| CursorBench | 39.6% | — |
| LMArena WebDev | 1584 | — |
| FrontierSWE | 19.6% | — |
| SciCode | 56.6% | — |
| WeirdML | 84.8% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
| ALE-Bench | 1,270 | — |
Agentic & Tool Use Gemini 3.8 Flash leads
Gemini 3.8 Flash: 41.8 (#21), Phi-4: 22.8 (#128)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| APEX-Agents | 64.3% | — |
| Berkeley Function Calling Leaderboard | — | 28.8% |
| Remote Labor Index | 5.8% | — |
| BALROG | — | 11.6% |
| GDP.pdf | 23.4% | — |
| Vending-Bench 2 | 5,094 | — |
Reasoning Gemini 3.8 Flash leads
Gemini 3.8 Flash: 76.9 (#5), Phi-4: 17.7 (#291)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| Chess Puzzles | 61% | 1% |
| LMArena Hard Prompts | 1508 | 1220 |
| Epoch Capabilities Index | 156.71 | 130.42 |
| ARC-AGI-2 | 89.2% | — |
| NYT Connections (extended) | 97.4% | — |
| ARC-AGI-1 | 98.5% | — |
| CritPt | 18.3% | — |
| LiveBench Reasoning | — | 47.8% |
| Mystery Game Puzzles | 47% | — |
| DTBench | 95.7% | — |
| LiveBench Data Analysis | — | 45.2% |
| LMCA | 52.9% | — |
| Surface Evolver Bench | 76.9% | — |
| LiveBench | — | 41.6% |
Math Gemini 3.8 Flash leads
Gemini 3.8 Flash: 65.3 (#28), Phi-4: 20.8 (#285)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.9% | 13.8% |
| LMArena Math | 1528 | 1246 |
| FrontierMath (Tiers 1-3) | 68.4% | — |
| FrontierMath Tier 4 | 22% | — |
| ProofBench | 48% | — |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |
Knowledge Gemini 3.8 Flash leads
Gemini 3.8 Flash: 74.8 (#2), Phi-4: 32.6 (#209)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| GPQA Diamond | 95.4% | 56.1% |
| LMArena Expert | 1524 | 1203 |
| Humanity's Last Exam | 44.5% | — |
| SimpleQA Verified | 69.7% | — |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| MMLU | — | 84.8% |
Multimodal Not comparable
Gemini 3.8 Flash: 40.7 (#45), Phi-4: —
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| LMArena Vision | 1314 | — |
| Blueprint-Bench 2 | 38.6% | — |
| Furniture Assembly | 31.7% | — |
Multilingual Gemini 3.8 Flash leads
Gemini 3.8 Flash: 58.0 (#5), Phi-4: 37.2 (#237)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| LMArena Non-English | 1491 | 1197 |
| LMArena Chinese | 1554 | 1212 |
| LMArena French | 1498 | 1224 |
| LMArena German | 1493 | 1222 |
| LMArena Japanese | 1502 | 1158 |
| LMArena Korean | 1459 | 1151 |
| LMArena Russian | 1515 | 1209 |
| LMArena Spanish | 1485 | 1234 |
Instruction Following Gemini 3.8 Flash leads
Gemini 3.8 Flash: 78.0 (#13), Phi-4: 60.4 (#251)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1490 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
Long Context Gemini 3.8 Flash leads
Gemini 3.8 Flash: 46.3 (#24), Phi-4: 36.9 (#226)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1508 | 1217 |
Writing & Preference Gemini 3.8 Flash leads
Gemini 3.8 Flash: 72.2 (#15), Phi-4: 40.5 (#244)
| Benchmark | Gemini 3.8 Flash | Phi-4 |
|---|---|---|
| LMArena Text | 1499 | 1217 |
| LMArena Creative Writing | 1492 | 1182 |
| LMArena Multi-Turn | 1501 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| EQ-Bench Creative Writing | 1748 | — |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is Gemini 3.8 Flash better than Phi-4?
Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 31.2 on the Noometry Index. Phi-4 costs 17× less per token, which makes it the better buy when Gemini 3.8 Flash's lead doesn't matter for your workload.
Which is cheaper, Gemini 3.8 Flash or Phi-4?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Gemini 3.8 Flash lists at $0.75 and $3.75.
Is Gemini 3.8 Flash or Phi-4 better for coding?
Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
Gemini 3.8 Flash does, with 1.05M tokens against 128K.
How many benchmarks do Gemini 3.8 Flash and Phi-4 share?
21 benchmarks have published results for both models. Gemini 3.8 Flash has 50 scored results on Noometry and Phi-4 has 37.