# GPT-5 Mini vs Llama 3.2 3B

> GPT-5 Mini is the stronger model overall, scoring 41.7 to 28.4 on the Noometry Index. Llama 3.2 3B costs 5.7× less per token, which makes it the better buy when GPT-5 Mini's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-mini-vs-llama-3-2-3b
- Last updated: 2026-10-10
- Shared benchmarks: 15

## Summary

- They share 15 benchmarks with published results for both. GPT-5 Mini scores higher in 9 categories and Llama 3.2 3B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-5 Mini leads 55.4 to 24.9.
- The biggest single-benchmark swing is Berkeley Function Calling Leaderboard: 55.5% for GPT-5 Mini and 21.9% for Llama 3.2 3B.
- Llama 3.2 3B is cheaper at $0.05 / $0.33 per million input/output tokens, against $0.25 / $2 for GPT-5 Mini.
- GPT-5 Mini accepts more context: 400K tokens versus 131K.
- Llama 3.2 3B has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 41.7 | 28.4 |
| Rank | 115 | 318 |
| Context | 400K | 131K |
| Input $/M | $0.25 | $0.05 |
| Output $/M | $2 | $0.33 |
| Weights | Proprietary | Open |

## Coding

- GPT-5 Mini: 40.0 (#134)
- Llama 3.2 3B: 27.1 (#314)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Coding | 1406 | 1098 |
| SWE-bench Verified | 64.7% | — |
| SWE-bench Verified (bash only) | 59.8% | — |
| SWE-bench Multilingual | 39.7% | — |
| SciCode | 39.2% | — |
| WeirdML | 52.7% | — |
| BigCodeBench Instruct | — | 23.4% |
| BigCodeBench Complete | — | 28.3% |
| ALE-Bench | 799.77 | — |
| AlgoTune | 1.38 | — |

## Agentic & Tool Use

- GPT-5 Mini: 31.0 (#64)
- Llama 3.2 3B: 19.9 (#135)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 55.5% | 21.9% |
| Terminal-Bench | 34.8% | — |
| BALROG | — | 10.1% |
| Vending-Bench 2 | -31.18 | — |

## Reasoning

- GPT-5 Mini: 24.1 (#153)
- Llama 3.2 3B: 20.1 (#225)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Hard Prompts | 1380 | 1095 |
| ARC-AGI-2 | 4.4% | — |
| Kagi LLM Benchmark | 70.3% | — |
| ARC-AGI-1 | 54.3% | — |
| CritPt | 0% | — |
| Chess Puzzles | 30% | — |
| EnigmaEval | 8.2% | — |
| Mystery Game Puzzles | 10% | — |
| DTBench | 80.5% | — |
| LMCA | 34.2% | — |
| Epoch Capabilities Index | 145.51 | — |
| ForecastBench | 61 | — |

## Math

- GPT-5 Mini: 46.3 (#59)
- Llama 3.2 3B: 29.9 (#213)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Math | 1377 | 1126 |
| FrontierMath (Tiers 1-3) | 46.7% | — |
| FrontierMath Tier 4 | 12.2% | — |
| OTIS Mock AIME 2024-2025 | 86.7% | — |
| ProofBench | 9% | — |
| Omni-MATH | 72.2% | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 27.2% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |

## Knowledge

- GPT-5 Mini: 45.5 (#80)
- Llama 3.2 3B: 29.1 (#232)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Expert | 1377 | 1090 |
| GPQA Diamond | 75% | — |
| Humanity's Last Exam | 19.4% | — |
| SimpleQA Verified | 21.6% | — |
| MMLU-Pro | 83.5% | — |
| Confabulations | 13.3% | — |
| Vectara Hallucination Rate | 12.9% | — |
| GPQA (HELM) | 75.6% | — |

## Multimodal

- GPT-5 Mini: 35.0 (#80)
- Llama 3.2 3B: —

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Vision | 1202 | — |
| VPCT | 40.2% | — |

## Multilingual

- GPT-5 Mini: 49.1 (#131)
- Llama 3.2 3B: 26.4 (#276)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Non-English | 1363 | 1019 |
| LMArena Chinese | 1387 | 1017 |
| LMArena German | 1366 | 1056 |
| LMArena Russian | 1362 | 949 |
| LMArena French | 1385 | — |
| LMArena Japanese | 1342 | — |
| LMArena Korean | 1306 | — |
| LMArena Spanish | 1355 | — |

## Instruction Following

- GPT-5 Mini: 76.4 (#43)
- Llama 3.2 3B: 56.3 (#270)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Instruction Following | 1357 | 1089 |
| IFEval | 92.7% | — |

## Long Context

- GPT-5 Mini: 41.9 (#127)
- Llama 3.2 3B: 33.4 (#256)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Longer Query | 1355 | 1101 |
| Fiction.LiveBench | 69.4% | — |

## Writing & Preference

- GPT-5 Mini: 55.4 (#143)
- Llama 3.2 3B: 24.9 (#302)

| Benchmark | GPT-5 Mini | Llama 3.2 3B |
|---|---|---|
| LMArena Text | 1373 | 1110 |
| LMArena Creative Writing | 1325 | 1094 |
| EQ-Bench Creative Writing | 1313 | 595 |
| LMArena Multi-Turn | 1363 | 1105 |
| Short-Story Creative Writing | 83.1% | — |
| WildBench | 85.5% | — |

## FAQ

### Is GPT-5 Mini better than Llama 3.2 3B?

GPT-5 Mini is the stronger model overall, scoring 41.7 to 28.4 on the Noometry Index. Llama 3.2 3B costs 5.7× less per token, which makes it the better buy when GPT-5 Mini's lead doesn't matter for your workload.

### Which is cheaper, GPT-5 Mini or Llama 3.2 3B?

Llama 3.2 3B is cheaper. It lists at $0.05 per million input tokens and $0.33 per million output tokens; GPT-5 Mini lists at $0.25 and $2.

### Is GPT-5 Mini or Llama 3.2 3B better for coding?

GPT-5 Mini scores higher on coding benchmarks: 40.0 versus 27.1 in the Noometry coding category.

### Which has the bigger context window?

GPT-5 Mini does, with 400K tokens against 131K.

### How many benchmarks do GPT-5 Mini and Llama 3.2 3B share?

15 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and Llama 3.2 3B has 18.