Model comparison
Llama-3.3-70B-Instruct vs Nova 2 Lite
Nova 2 Lite is the stronger model overall, scoring 39.7 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 5.5× less per token, which makes it the better buy when Nova 2 Lite's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
MetaSummary
- They share 19 benchmarks with published results for both. Llama-3.3-70B-Instruct scores higher in 2 categories and Nova 2 Lite in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Nova 2 Lite leads 37.5 to 15.3.
- Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $0.30 / $2.50 for Nova 2 Lite.
- Nova 2 Lite accepts more context: 1M 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 | Nova 2 Lite | |
|---|---|---|
| Provider | Meta | Amazon |
| Noometry Index | 30.6 | 39.7 |
| Released | 2024-12-06 | 2025-12-01 |
| Weights | Open | Proprietary |
| Context window | 128K | 1M |
| Max output | 4K | 64K |
| Input $ / M tokens | $0.10 | $0.30 |
| Output $ / M tokens | $0.32 | $2.50 |
| Results tracked | 43 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Nova 2 Lite leads
Llama-3.3-70B-Instruct: 31.0 (#290), Nova 2 Lite: 40.7 (#134)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| LMArena Coding | 1268 | 1385 |
| SciCode | 26% | — |
| WeirdML | 14.4% | — |
| BigCodeBench Instruct | 46.9% | — |
| LiveBench Coding | 36.6% | — |
| BigCodeBench Complete | 57.5% | — |
Agentic & Tool Use Llama-3.3-70B-Instruct leads
Llama-3.3-70B-Instruct: 25.8 (#105), Nova 2 Lite: 24.1 (#121)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| Berkeley Function Calling Leaderboard | 31.9% | 27.1% |
| BALROG | 23% | — |
Reasoning Nova 2 Lite leads
Llama-3.3-70B-Instruct: 14.1 (#327), Nova 2 Lite: 27.5 (#118)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| LMArena Hard Prompts | 1257 | 1364 |
| SimpleBench | 19.9% | — |
| CritPt | 0% | — |
| LiveBench Reasoning | 50.8% | — |
| DTBench | 59.5% | — |
| LiveBench Data Analysis | 49.5% | — |
| LMCA | 17.5% | — |
| Epoch Capabilities Index | 127.33 | — |
| ForecastBench | 58.6 | — |
| LiveBench | 50.2% | — |
Math Nova 2 Lite leads
Llama-3.3-70B-Instruct: 15.3 (#298), Nova 2 Lite: 37.5 (#156)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| LMArena Math | 1267 | 1359 |
| OTIS Mock AIME 2024-2025 | 5.1% | — |
| LiveBench Math | 42.2% | — |
| MATH Level 5 | 41.6% | — |
Knowledge Nova 2 Lite leads
Llama-3.3-70B-Instruct: 30.6 (#226), Nova 2 Lite: 43.0 (#94)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| Vectara Hallucination Rate | 4.1% | 5.1% |
| LMArena Expert | 1225 | 1358 |
| GPQA Diamond | 47.4% | — |
| Confabulations | 22.8% | — |
| MMLU | 86.3% | — |
Multilingual Nova 2 Lite leads
Llama-3.3-70B-Instruct: 39.9 (#220), Nova 2 Lite: 47.1 (#153)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| LMArena Non-English | 1236 | 1337 |
| LMArena Chinese | 1217 | 1364 |
| LMArena French | 1281 | 1381 |
| LMArena German | 1251 | 1343 |
| LMArena Japanese | 1150 | 1271 |
| LMArena Korean | 1143 | 1284 |
| LMArena Russian | 1252 | 1343 |
| LMArena Spanish | 1270 | 1373 |
Instruction Following Too close to call
Llama-3.3-70B-Instruct: 71.1 (#157), Nova 2 Lite: 70.5 (#161)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| LMArena Instruction Following | 1242 | 1335 |
| LiveBench Instruction Following | 82.7% | — |
Long Context Nova 2 Lite leads
Llama-3.3-70B-Instruct: 26.4 (#295), Nova 2 Lite: 40.6 (#150)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| LMArena Longer Query | 1256 | 1335 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference Nova 2 Lite leads
Llama-3.3-70B-Instruct: 47.6 (#207), Nova 2 Lite: 53.9 (#154)
| Benchmark | Llama-3.3-70B-Instruct | Nova 2 Lite |
|---|---|---|
| LMArena Text | 1274 | 1362 |
| LMArena Creative Writing | 1250 | 1291 |
| LMArena Multi-Turn | 1280 | 1338 |
| LiveBench Language | 39.2% | — |
Frequently asked questions
Is Llama-3.3-70B-Instruct better than Nova 2 Lite?
Nova 2 Lite is the stronger model overall, scoring 39.7 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 5.5× less per token, which makes it the better buy when Nova 2 Lite's lead doesn't matter for your workload.
Which is cheaper, Llama-3.3-70B-Instruct or Nova 2 Lite?
Llama-3.3-70B-Instruct is cheaper. It lists at $0.10 per million input tokens and $0.32 per million output tokens; Nova 2 Lite lists at $0.30 and $2.50.
Is Llama-3.3-70B-Instruct or Nova 2 Lite better for coding?
Nova 2 Lite scores higher on coding benchmarks: 40.7 versus 31.0 in the Noometry coding category.
Which has the bigger context window?
Nova 2 Lite does, with 1M tokens against 128K.
How many benchmarks do Llama-3.3-70B-Instruct and Nova 2 Lite share?
19 benchmarks have published results for both models. Llama-3.3-70B-Instruct has 43 scored results on Noometry and Nova 2 Lite has 19.