Model comparison
Llama 3.2 1B vs o1
o1 is the stronger model overall, scoring 40.9 to 20.1 on the Noometry Index. Llama 3.2 1B costs 372× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Last verified . 17 shared benchmarks.
MetaSummary
- They share 17 benchmarks with published results for both. Llama 3.2 1B scores higher in 0 categories and o1 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where o1 leads 41.5 to 7.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 0.6% for Llama 3.2 1B and 73.3% for o1.
- Llama 3.2 1B is cheaper at $0.027 / $0.20 per million input/output tokens, against $15 / $60 for o1.
- o1 accepts more context: 200K tokens versus 60K.
- Llama 3.2 1B has downloadable open weights; the other is API-only.
Side by side
| Llama 3.2 1B | o1 | |
|---|---|---|
| Provider | Meta | OpenAI |
| Noometry Index | 20.1 | 40.9 |
| Released | 2024-09-24 | 2024-09-12 |
| Weights | Open | Proprietary |
| Context window | 60K | 200K |
| Max output | 54K | 100K |
| Input $ / M tokens | $0.027 | $15 |
| Output $ / M tokens | $0.20 | $60 |
| Results tracked | 22 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o1 leads
Llama 3.2 1B: 21.1 (#338), o1: 46.1 (#70)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| LMArena Coding | 1070 | 1367 |
| Aider Polyglot | — | 61.7% |
| WeirdML | — | 47.6% |
| BigCodeBench Instruct | 8.2% | — |
| LiveBench Coding | — | 69.7% |
| BigCodeBench Complete | 11.3% | — |
| CadEval | — | 56% |
| HumanEval+ | — | 89% |
| MBPP+ | — | 80.2% |
Agentic & Tool Use o1 leads
Llama 3.2 1B: 14.6 (#150), o1: 24.6 (#117)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 10.8% | — |
| Cybench | — | 10% |
| BALROG | 6.6% | — |
| METR Time Horizons | — | 51.1% |
Reasoning o1 leads
Llama 3.2 1B: 16.2 (#308), o1: 27.9 (#111)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| Chess Puzzles | 0% | 15% |
| LMArena Hard Prompts | 1044 | 1371 |
| Epoch Capabilities Index | 101.99 | 141.91 |
| SimpleBench | — | 41.7% |
| ARC-AGI-1 | — | 30.7% |
| EnigmaEval | — | 5.7% |
| LiveBench Reasoning | — | 91.6% |
| DTBench | — | 74.7% |
| LiveBench Data Analysis | — | 65.5% |
| LMCA | — | 22.3% |
| LiveBench | — | 75.7% |
Math o1 leads
Llama 3.2 1B: 10.4 (#313), o1: 36.1 (#175)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 0.6% | 73.3% |
| LMArena Math | 1086 | 1388 |
| FrontierMath (Tiers 1-3) | — | 14.7% |
| LiveBench Math | — | 80.3% |
| MATH Level 5 | — | 94.7% |
| FrontierMath (Feb 2025 set) | — | 9.3% |
Knowledge o1 leads
Llama 3.2 1B: 7.2 (#312), o1: 41.5 (#110)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| GPQA Diamond | 23.9% | 76.8% |
| LMArena Expert | 1007 | 1361 |
| Humanity's Last Exam | — | 8% |
| SimpleQA Verified | — | 41.1% |
| Confabulations | — | 11.7% |
Multimodal Not comparable
Llama 3.2 1B: —, o1: 34.2 (#93)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| LMArena Vision | — | 1168 |
| GeoBench | — | 80% |
| VPCT | — | 37% |
| SpatialViz-Bench | — | 41.4% |
Multilingual o1 leads
Llama 3.2 1B: 23.8 (#292), o1: 48.6 (#142)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| LMArena Non-English | 973 | 1358 |
| LMArena Chinese | 959 | 1394 |
| LMArena German | 1014 | 1337 |
| LMArena Russian | 941 | 1356 |
| LMArena French | — | 1344 |
| LMArena Japanese | — | 1346 |
| LMArena Korean | — | 1396 |
| LMArena Spanish | — | 1345 |
Instruction Following o1 leads
Llama 3.2 1B: 52.4 (#290), o1: 74.8 (#86)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| LMArena Instruction Following | 1031 | 1367 |
| LiveBench Instruction Following | — | 81.5% |
Long Context o1 leads
Llama 3.2 1B: 31.9 (#274), o1: 50.3 (#9)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| LMArena Longer Query | 1050 | 1378 |
| Fiction.LiveBench | — | 83.3% |
Writing & Preference o1 leads
Llama 3.2 1B: 21.3 (#310), o1: 55.6 (#144)
| Benchmark | Llama 3.2 1B | o1 |
|---|---|---|
| LMArena Text | 1055 | 1366 |
| LMArena Creative Writing | 1033 | 1348 |
| LMArena Multi-Turn | 1030 | 1369 |
| Short-Story Creative Writing | — | 70.2% |
| EQ-Bench Creative Writing | 200 | — |
| LiveBench Language | — | 65.4% |
Frequently asked questions
Is Llama 3.2 1B better than o1?
o1 is the stronger model overall, scoring 40.9 to 20.1 on the Noometry Index. Llama 3.2 1B costs 372× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Which is cheaper, Llama 3.2 1B or o1?
Llama 3.2 1B is cheaper. It lists at $0.027 per million input tokens and $0.20 per million output tokens; o1 lists at $15 and $60.
Is Llama 3.2 1B or o1 better for coding?
o1 scores higher on coding benchmarks: 46.1 versus 21.1 in the Noometry coding category.
Which has the bigger context window?
o1 does, with 200K tokens against 60K.
How many benchmarks do Llama 3.2 1B and o1 share?
17 benchmarks have published results for both models. Llama 3.2 1B has 22 scored results on Noometry and o1 has 52.