Model comparison
Llama-3.3-70B-Instruct vs o1
o1 is the stronger model overall, scoring 40.9 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 169× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Last verified . 34 shared benchmarks.
MetaSummary
- They share 34 benchmarks with published results for both. Llama-3.3-70B-Instruct scores higher in 1 category and o1 in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in long context, where o1 leads 50.3 to 26.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 5.1% for Llama-3.3-70B-Instruct and 73.3% for o1.
- Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $15 / $60 for o1.
- o1 accepts more context: 200K tokens versus 128K.
- Llama-3.3-70B-Instruct has downloadable open weights; the other is API-only.
Side by side
| Llama-3.3-70B-Instruct | o1 | |
|---|---|---|
| Provider | Meta | OpenAI |
| Noometry Index | 30.6 | 40.9 |
| Released | 2024-12-06 | 2024-09-12 |
| Weights | Open | Proprietary |
| Context window | 128K | 200K |
| Max output | 4K | 100K |
| Input $ / M tokens | $0.10 | $15 |
| Output $ / M tokens | $0.32 | $60 |
| Results tracked | 43 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o1 leads
Llama-3.3-70B-Instruct: 31.0 (#290), o1: 46.1 (#70)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| WeirdML | 14.4% | 47.6% |
| LiveBench Coding | 36.6% | 69.7% |
| LMArena Coding | 1268 | 1367 |
| Aider Polyglot | — | 61.7% |
| SciCode | 26% | — |
| BigCodeBench Instruct | 46.9% | — |
| BigCodeBench Complete | 57.5% | — |
| CadEval | — | 56% |
| HumanEval+ | — | 89% |
| MBPP+ | — | 80.2% |
Agentic & Tool Use Llama-3.3-70B-Instruct leads
Llama-3.3-70B-Instruct: 25.8 (#105), o1: 24.6 (#117)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 31.9% | — |
| Cybench | — | 10% |
| BALROG | 23% | — |
| METR Time Horizons | — | 51.1% |
Reasoning o1 leads
Llama-3.3-70B-Instruct: 14.1 (#327), o1: 27.9 (#111)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| SimpleBench | 19.9% | 41.7% |
| LiveBench Reasoning | 50.8% | 91.6% |
| LMArena Hard Prompts | 1257 | 1371 |
| DTBench | 59.5% | 74.7% |
| LiveBench Data Analysis | 49.5% | 65.5% |
| LMCA | 17.5% | 22.3% |
| Epoch Capabilities Index | 127.33 | 141.91 |
| LiveBench | 50.2% | 75.7% |
| ARC-AGI-1 | — | 30.7% |
| CritPt | 0% | — |
| Chess Puzzles | — | 15% |
| EnigmaEval | — | 5.7% |
| ForecastBench | 58.6 | — |
Math o1 leads
Llama-3.3-70B-Instruct: 15.3 (#298), o1: 36.1 (#175)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 5.1% | 73.3% |
| LiveBench Math | 42.2% | 80.3% |
| LMArena Math | 1267 | 1388 |
| MATH Level 5 | 41.6% | 94.7% |
| FrontierMath (Tiers 1-3) | — | 14.7% |
| FrontierMath (Feb 2025 set) | — | 9.3% |
Knowledge o1 leads
Llama-3.3-70B-Instruct: 30.6 (#226), o1: 41.5 (#110)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| GPQA Diamond | 47.4% | 76.8% |
| Confabulations | 22.8% | 11.7% |
| LMArena Expert | 1225 | 1361 |
| Humanity's Last Exam | — | 8% |
| SimpleQA Verified | — | 41.1% |
| Vectara Hallucination Rate | 4.1% | — |
| MMLU | 86.3% | — |
Multimodal Not comparable
Llama-3.3-70B-Instruct: —, o1: 34.2 (#93)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| LMArena Vision | — | 1168 |
| GeoBench | — | 80% |
| VPCT | — | 37% |
| SpatialViz-Bench | — | 41.4% |
Multilingual o1 leads
Llama-3.3-70B-Instruct: 39.9 (#220), o1: 48.6 (#142)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| LMArena Non-English | 1236 | 1358 |
| LMArena Chinese | 1217 | 1394 |
| LMArena French | 1281 | 1344 |
| LMArena German | 1251 | 1337 |
| LMArena Japanese | 1150 | 1346 |
| LMArena Korean | 1143 | 1396 |
| LMArena Russian | 1252 | 1356 |
| LMArena Spanish | 1270 | 1345 |
Instruction Following o1 leads
Llama-3.3-70B-Instruct: 71.1 (#157), o1: 74.8 (#86)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| LiveBench Instruction Following | 82.7% | 81.5% |
| LMArena Instruction Following | 1242 | 1367 |
Long Context o1 leads
Llama-3.3-70B-Instruct: 26.4 (#295), o1: 50.3 (#9)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| Fiction.LiveBench | 33.3% | 83.3% |
| LMArena Longer Query | 1256 | 1378 |
Writing & Preference o1 leads
Llama-3.3-70B-Instruct: 47.6 (#207), o1: 55.6 (#144)
| Benchmark | Llama-3.3-70B-Instruct | o1 |
|---|---|---|
| LMArena Text | 1274 | 1366 |
| LMArena Creative Writing | 1250 | 1348 |
| LMArena Multi-Turn | 1280 | 1369 |
| LiveBench Language | 39.2% | 65.4% |
| Short-Story Creative Writing | — | 70.2% |
Frequently asked questions
Is Llama-3.3-70B-Instruct better than o1?
o1 is the stronger model overall, scoring 40.9 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 169× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Which is cheaper, Llama-3.3-70B-Instruct or o1?
Llama-3.3-70B-Instruct is cheaper. It lists at $0.10 per million input tokens and $0.32 per million output tokens; o1 lists at $15 and $60.
Is Llama-3.3-70B-Instruct or o1 better for coding?
o1 scores higher on coding benchmarks: 46.1 versus 31.0 in the Noometry coding category.
Which has the bigger context window?
o1 does, with 200K tokens against 128K.
How many benchmarks do Llama-3.3-70B-Instruct and o1 share?
34 benchmarks have published results for both models. Llama-3.3-70B-Instruct has 43 scored results on Noometry and o1 has 52.