Model comparison
GPT-4.5 vs Llama-3.3-70B-Instruct
GPT-4.5 is the stronger model overall, scoring 37.2 to 30.6 on the Noometry Index.
Last verified . 32 shared benchmarks.
MetaSummary
- They share 32 benchmarks with published results for both. GPT-4.5 scores higher in 8 categories and Llama-3.3-70B-Instruct in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-4.5 leads 32.6 to 15.3.
- The biggest single-benchmark swing is LiveBench Coding: 75.2% for GPT-4.5 and 36.6% for Llama-3.3-70B-Instruct.
- Llama-3.3-70B-Instruct has downloadable open weights; the other is API-only.
Side by side
| GPT-4.5 | Llama-3.3-70B-Instruct | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 37.2 | 30.6 |
| Released | 2025-02-27 | 2024-12-06 |
| Weights | Proprietary | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.10 |
| Output $ / M tokens | — | $0.32 |
| Results tracked | 42 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-4.5 leads
GPT-4.5: 42.2 (#109), Llama-3.3-70B-Instruct: 31.0 (#290)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| WeirdML | 39.4% | 14.4% |
| LiveBench Coding | 75.2% | 36.6% |
| LMArena Coding | 1396 | 1268 |
| Aider Polyglot | 44.9% | — |
| SciCode | — | 26% |
| BigCodeBench Instruct | — | 46.9% |
| BigCodeBench Complete | — | 57.5% |
Agentic & Tool Use GPT-4.5 leads
GPT-4.5: 27.9 (#97), Llama-3.3-70B-Instruct: 25.8 (#105)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 31.9% |
| Cybench | 17.5% | — |
| BALROG | — | 23% |
Reasoning Too close to call
GPT-4.5: 13.9 (#330), Llama-3.3-70B-Instruct: 14.1 (#327)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| SimpleBench | 34.5% | 19.9% |
| LiveBench Reasoning | 71.1% | 50.8% |
| LMArena Hard Prompts | 1403 | 1257 |
| LiveBench Data Analysis | 64.3% | 49.5% |
| Epoch Capabilities Index | 136.74 | 127.33 |
| ForecastBench | 61.7 | 58.6 |
| LiveBench | 69% | 50.2% |
| ARC-AGI-2 | 0.8% | — |
| ARC-AGI-1 | 10.3% | — |
| CritPt | — | 0% |
| EnigmaEval | 3.2% | — |
| DTBench | — | 59.5% |
| LMCA | — | 17.5% |
Math GPT-4.5 leads
GPT-4.5: 32.6 (#211), Llama-3.3-70B-Instruct: 15.3 (#298)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 5.1% |
| LiveBench Math | 69.3% | 42.2% |
| LMArena Math | 1412 | 1267 |
| MATH Level 5 | 78.6% | 41.6% |
Knowledge GPT-4.5 leads
GPT-4.5: 32.5 (#211), Llama-3.3-70B-Instruct: 30.6 (#226)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| GPQA Diamond | 68.7% | 47.4% |
| Confabulations | 13.6% | 22.8% |
| LMArena Expert | 1394 | 1225 |
| Humanity's Last Exam | 5.4% | — |
| Vectara Hallucination Rate | — | 4.1% |
| MMLU | — | 86.3% |
Multimodal Not comparable
GPT-4.5: 37.6 (#71), Llama-3.3-70B-Instruct: —
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Vision | 1195 | — |
| VPCT | 45% | — |
Multilingual GPT-4.5 leads
GPT-4.5: 52.5 (#83), Llama-3.3-70B-Instruct: 39.9 (#220)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1413 | 1236 |
| LMArena Chinese | 1421 | 1217 |
| LMArena French | 1418 | 1281 |
| LMArena German | 1457 | 1251 |
| LMArena Japanese | 1416 | 1150 |
| LMArena Korean | 1392 | 1143 |
| LMArena Russian | 1419 | 1252 |
| LMArena Spanish | — | 1270 |
Instruction Following GPT-4.5 leads
GPT-4.5: 72.6 (#134), Llama-3.3-70B-Instruct: 71.1 (#157)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| LiveBench Instruction Following | 72.3% | 82.7% |
| LMArena Instruction Following | 1404 | 1242 |
Long Context GPT-4.5 leads
GPT-4.5: 40.4 (#155), Llama-3.3-70B-Instruct: 26.4 (#295)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| Fiction.LiveBench | 63.9% | 33.3% |
| LMArena Longer Query | 1406 | 1256 |
Writing & Preference GPT-4.5 leads
GPT-4.5: 56.9 (#134), Llama-3.3-70B-Instruct: 47.6 (#207)
| Benchmark | GPT-4.5 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1417 | 1274 |
| LMArena Creative Writing | 1394 | 1250 |
| LMArena Multi-Turn | 1444 | 1280 |
| LiveBench Language | 61.5% | 39.2% |
| Short-Story Creative Writing | 75.6% | — |
| EQ-Bench Creative Writing | 1258 | — |
Frequently asked questions
Is GPT-4.5 better than Llama-3.3-70B-Instruct?
GPT-4.5 is the stronger model overall, scoring 37.2 to 30.6 on the Noometry Index.
Is GPT-4.5 or Llama-3.3-70B-Instruct better for coding?
GPT-4.5 scores higher on coding benchmarks: 42.2 versus 31.0 in the Noometry coding category.
How many benchmarks do GPT-4.5 and Llama-3.3-70B-Instruct share?
32 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Llama-3.3-70B-Instruct has 43.