Model comparison
GPT-4o mini vs Llama 3.2 90B
Llama 3.2 90B is the stronger model overall, scoring 27.1 to 25.4 on the Noometry Index.
Last verified . 8 shared benchmarks.
MetaSummary
- They share 8 benchmarks with published results for both. GPT-4o mini scores higher in 1 category and Llama 3.2 90B in 4 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Llama 3.2 90B leads 20.9 to 8.7.
- The biggest single-benchmark swing is MATH Level 5: 52.6% for GPT-4o mini and 39.4% for Llama 3.2 90B.
- Llama 3.2 90B has downloadable open weights; the other is API-only.
Side by side
| GPT-4o mini | Llama 3.2 90B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 25.4 | 27.1 |
| Released | 2024-07-18 | 2024-09-24 |
| Weights | Proprietary | Open |
| Context window | 128K | — |
| Max output | 16K | — |
| Input $ / M tokens | $0.15 | — |
| Output $ / M tokens | $0.60 | — |
| Results tracked | 60 | 9 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
GPT-4o mini: 21.6 (#330), Llama 3.2 90B: —
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| Aider Polyglot | 3.6% | — |
| WeirdML | 11.8% | — |
| BigCodeBench Instruct | 46.1% | — |
| LiveBench Coding | 43.1% | — |
| LMArena Coding | 1290 | — |
| BigCodeBench Complete | 57.4% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use Llama 3.2 90B leads
GPT-4o mini: 27.2 (#95), Llama 3.2 90B: 29.7 (#73)
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| BALROG | 17.4% | 27.3% |
Reasoning Llama 3.2 90B leads
GPT-4o mini: 8.7 (#343), Llama 3.2 90B: 20.9 (#207)
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| Epoch Capabilities Index | 126.56 | 125.5 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 10.7% | — |
| Kagi LLM Benchmark | 28.8% | — |
| Chess Puzzles | 0% | — |
| EnigmaEval | — | 0.4% |
| LiveBench Reasoning | 32.8% | — |
| LMArena Hard Prompts | 1267 | — |
| Mystery Game Puzzles | 12% | — |
| DTBench | 54.4% | — |
| LiveBench Data Analysis | 50% | — |
| LMCA | 10.4% | — |
| LiveBench | 41.3% | — |
| PIQA | 88.7% | — |
Math Too close to call
GPT-4o mini: 9.9 (#307), Llama 3.2 90B: 10.4 (#302)
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 6.9% | 2.6% |
| MATH Level 5 | 52.6% | 39.4% |
| FrontierMath (Tiers 1-3) | 0.7% | — |
| Omni-MATH | 28% | — |
| LiveBench Math | 36.3% | — |
| LMArena Math | 1267 | — |
| GSM8K | 91.3% | — |
Knowledge Llama 3.2 90B leads
GPT-4o mini: 17.7 (#279), Llama 3.2 90B: 21.7 (#269)
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| GPQA Diamond | 37.7% | 41% |
| MMLU | 81.8% | 80.3% |
| SimpleQA Verified | 8.3% | — |
| MMLU-Pro | 60.3% | — |
| Confabulations | 37.2% | — |
| GPQA (HELM) | 36.8% | — |
| LMArena Expert | 1235 | — |
| BoolQ | 88.7% | — |
Multimodal Too close to call
GPT-4o mini: 25.3 (#117), Llama 3.2 90B: 24.9 (#119)
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| LMArena Vision | 1066 | 1000 |
| GeoBench | 64% | 52% |
| Video-MME | 64.8% | — |
| VPCT | 34% | — |
Multilingual Not comparable
GPT-4o mini: 42.2 (#194), Llama 3.2 90B: —
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| LMArena Non-English | 1266 | — |
| LMArena Chinese | 1264 | — |
| LMArena French | 1296 | — |
| LMArena German | 1272 | — |
| LMArena Japanese | 1216 | — |
| LMArena Korean | 1195 | — |
| LMArena Russian | 1275 | — |
| LMArena Spanish | 1276 | — |
Instruction Following Not comparable
GPT-4o mini: 62.1 (#234), Llama 3.2 90B: —
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| LiveBench Instruction Following | 56.8% | — |
| IFEval | 78.2% | — |
| LMArena Instruction Following | 1259 | — |
Long Context Not comparable
GPT-4o mini: 39.1 (#181), Llama 3.2 90B: —
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| LMArena Longer Query | 1289 | — |
Writing & Preference Not comparable
GPT-4o mini: 39.7 (#243), Llama 3.2 90B: —
| Benchmark | GPT-4o mini | Llama 3.2 90B |
|---|---|---|
| LMArena Text | 1286 | — |
| LMArena Creative Writing | 1268 | — |
| Short-Story Creative Writing | 67.2% | — |
| EQ-Bench Creative Writing | 873 | — |
| WildBench | 79.1% | — |
| LMArena Multi-Turn | 1285 | — |
| LiveBench Language | 28.6% | — |
Frequently asked questions
Is GPT-4o mini better than Llama 3.2 90B?
Llama 3.2 90B is the stronger model overall, scoring 27.1 to 25.4 on the Noometry Index.
How many benchmarks do GPT-4o mini and Llama 3.2 90B share?
8 benchmarks have published results for both models. GPT-4o mini has 60 scored results on Noometry and Llama 3.2 90B has 9.