Model comparison
Llama-3.3-70B-Instruct vs Phi-4
Llama-3.3-70B-Instruct and Phi-4 score almost the same on the Noometry Index (30.6 vs 31.2), so choose on price, context window or the category you care about most.
Last verified . 35 shared benchmarks.
Meta
MicrosoftSummary
- They share 35 benchmarks with published results for both. Llama-3.3-70B-Instruct scores higher in 4 categories and Phi-4 in 5 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in instruction following, where Llama-3.3-70B-Instruct leads 71.1 to 60.4.
- The biggest single-benchmark swing is LiveBench Instruction Following: 82.7% for Llama-3.3-70B-Instruct and 58.4% for Phi-4.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.10 / $0.32 for Llama-3.3-70B-Instruct.
Side by side
| Llama-3.3-70B-Instruct | Phi-4 | |
|---|---|---|
| Provider | Meta | Microsoft |
| Noometry Index | 30.6 | 31.2 |
| Released | 2024-12-06 | 2024-12-11 |
| Weights | Open | Open |
| Context window | 128K | 128K |
| Max output | 4K | 4K |
| Input $ / M tokens | $0.10 | $0.07 |
| Output $ / M tokens | $0.32 | $0.14 |
| Results tracked | 43 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Phi-4 leads
Llama-3.3-70B-Instruct: 31.0 (#290), Phi-4: 34.4 (#239)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| BigCodeBench Instruct | 46.9% | 45.5% |
| LiveBench Coding | 36.6% | 30.7% |
| LMArena Coding | 1268 | 1231 |
| BigCodeBench Complete | 57.5% | 55.4% |
| SciCode | 26% | — |
| WeirdML | 14.4% | — |
Agentic & Tool Use Llama-3.3-70B-Instruct leads
Llama-3.3-70B-Instruct: 25.8 (#105), Phi-4: 22.8 (#128)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 31.9% | 28.8% |
| BALROG | 23% | 11.6% |
Reasoning Phi-4 leads
Llama-3.3-70B-Instruct: 14.1 (#327), Phi-4: 17.7 (#291)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| LiveBench Reasoning | 50.8% | 47.8% |
| LMArena Hard Prompts | 1257 | 1220 |
| LiveBench Data Analysis | 49.5% | 45.2% |
| Epoch Capabilities Index | 127.33 | 130.42 |
| LiveBench | 50.2% | 41.6% |
| SimpleBench | 19.9% | — |
| CritPt | 0% | — |
| Chess Puzzles | — | 1% |
| DTBench | 59.5% | — |
| LMCA | 17.5% | — |
| ForecastBench | 58.6 | — |
Math Phi-4 leads
Llama-3.3-70B-Instruct: 15.3 (#298), Phi-4: 20.8 (#285)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 5.1% | 13.8% |
| LiveBench Math | 42.2% | 42% |
| LMArena Math | 1267 | 1246 |
| MATH Level 5 | 41.6% | 64.9% |
Knowledge Phi-4 leads
Llama-3.3-70B-Instruct: 30.6 (#226), Phi-4: 32.6 (#209)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| GPQA Diamond | 47.4% | 56.1% |
| Confabulations | 22.8% | 29.4% |
| Vectara Hallucination Rate | 4.1% | 3.7% |
| LMArena Expert | 1225 | 1203 |
| MMLU | 86.3% | 84.8% |
Multilingual Llama-3.3-70B-Instruct leads
Llama-3.3-70B-Instruct: 39.9 (#220), Phi-4: 37.2 (#237)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| LMArena Non-English | 1236 | 1197 |
| LMArena Chinese | 1217 | 1212 |
| LMArena French | 1281 | 1224 |
| LMArena German | 1251 | 1222 |
| LMArena Japanese | 1150 | 1158 |
| LMArena Korean | 1143 | 1151 |
| LMArena Russian | 1252 | 1209 |
| LMArena Spanish | 1270 | 1234 |
Instruction Following Llama-3.3-70B-Instruct leads
Llama-3.3-70B-Instruct: 71.1 (#157), Phi-4: 60.4 (#251)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| LiveBench Instruction Following | 82.7% | 58.4% |
| LMArena Instruction Following | 1242 | 1201 |
Long Context Phi-4 leads
Llama-3.3-70B-Instruct: 26.4 (#295), Phi-4: 36.9 (#226)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1256 | 1217 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference Llama-3.3-70B-Instruct leads
Llama-3.3-70B-Instruct: 47.6 (#207), Phi-4: 40.5 (#244)
| Benchmark | Llama-3.3-70B-Instruct | Phi-4 |
|---|---|---|
| LMArena Text | 1274 | 1217 |
| LMArena Creative Writing | 1250 | 1182 |
| LMArena Multi-Turn | 1280 | 1206 |
| LiveBench Language | 39.2% | 25.6% |
| Short-Story Creative Writing | — | 62.6% |
Frequently asked questions
Is Llama-3.3-70B-Instruct better than Phi-4?
Llama-3.3-70B-Instruct and Phi-4 score almost the same on the Noometry Index (30.6 vs 31.2), so choose on price, context window or the category you care about most.
Which is cheaper, Llama-3.3-70B-Instruct or Phi-4?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Llama-3.3-70B-Instruct lists at $0.10 and $0.32.
Is Llama-3.3-70B-Instruct or Phi-4 better for coding?
Phi-4 scores higher on coding benchmarks: 34.4 versus 31.0 in the Noometry coding category.
Which has the bigger context window?
Both accept 128K tokens.
How many benchmarks do Llama-3.3-70B-Instruct and Phi-4 share?
35 benchmarks have published results for both models. Llama-3.3-70B-Instruct has 43 scored results on Noometry and Phi-4 has 37.