Model comparison
Nemotron 4 340b Instruct vs Olmo 3 32b Think
Olmo 3 32b Think is the stronger model overall, scoring 38.7 to 35.9 on the Noometry Index.
Last verified . 14 shared benchmarks.
NVIDIA
Allen Institute for AI (Ai2)Summary
- They share 14 benchmarks with published results for both. Nemotron 4 340b Instruct scores higher in 0 categories and Olmo 3 32b Think in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Olmo 3 32b Think leads 49.1 to 42.7.
Side by side
| Nemotron 4 340b Instruct | Olmo 3 32b Think | |
|---|---|---|
| Provider | NVIDIA | Allen Institute for AI (Ai2) |
| Noometry Index | 35.9 | 38.7 |
| Released | 2024-06-14 | — |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 17 | 14 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Olmo 3 32b Think leads
Nemotron 4 340b Instruct: 35.2 (#228), Olmo 3 32b Think: 38.6 (#172)
| Benchmark | Nemotron 4 340b Instruct | Olmo 3 32b Think |
|---|---|---|
| LMArena Coding | 1209 | 1319 |
Reasoning Olmo 3 32b Think leads
Nemotron 4 340b Instruct: 23.4 (#178), Olmo 3 32b Think: 25.9 (#140)
| Benchmark | Nemotron 4 340b Instruct | Olmo 3 32b Think |
|---|---|---|
| LMArena Hard Prompts | 1202 | 1302 |
Math Olmo 3 32b Think leads
Nemotron 4 340b Instruct: 34.3 (#195), Olmo 3 32b Think: 36.5 (#165)
| Benchmark | Nemotron 4 340b Instruct | Olmo 3 32b Think |
|---|---|---|
| LMArena Math | 1216 | 1316 |
Knowledge Olmo 3 32b Think leads
Nemotron 4 340b Instruct: 32.0 (#216), Olmo 3 32b Think: 35.0 (#190)
| Benchmark | Nemotron 4 340b Instruct | Olmo 3 32b Think |
|---|---|---|
| LMArena Expert | 1171 | 1273 |
Multilingual Olmo 3 32b Think leads
Nemotron 4 340b Instruct: 37.9 (#233), Olmo 3 32b Think: 41.2 (#210)
| Benchmark | Nemotron 4 340b Instruct | Olmo 3 32b Think |
|---|---|---|
| LMArena Non-English | 1206 | 1255 |
| LMArena Chinese | 1214 | 1300 |
| LMArena French | 1197 | 1291 |
| LMArena German | 1192 | 1290 |
| LMArena Russian | 1219 | 1254 |
| LMArena Japanese | 1128 | — |
| LMArena Korean | 1148 | — |
| LMArena Spanish | 1195 | — |
Instruction Following Olmo 3 32b Think leads
Nemotron 4 340b Instruct: 63.1 (#232), Olmo 3 32b Think: 67.2 (#198)
| Benchmark | Nemotron 4 340b Instruct | Olmo 3 32b Think |
|---|---|---|
| LMArena Instruction Following | 1203 | 1275 |
Long Context Olmo 3 32b Think leads
Nemotron 4 340b Instruct: 37.1 (#221), Olmo 3 32b Think: 39.4 (#182)
| Benchmark | Nemotron 4 340b Instruct | Olmo 3 32b Think |
|---|---|---|
| LMArena Longer Query | 1225 | 1296 |
Writing & Preference Olmo 3 32b Think leads
Nemotron 4 340b Instruct: 42.7 (#233), Olmo 3 32b Think: 49.1 (#193)
| Benchmark | Nemotron 4 340b Instruct | Olmo 3 32b Think |
|---|---|---|
| LMArena Text | 1225 | 1300 |
| LMArena Creative Writing | 1203 | 1256 |
| LMArena Multi-Turn | 1214 | 1290 |
Frequently asked questions
Is Nemotron 4 340b Instruct better than Olmo 3 32b Think?
Olmo 3 32b Think is the stronger model overall, scoring 38.7 to 35.9 on the Noometry Index.
Is Nemotron 4 340b Instruct or Olmo 3 32b Think better for coding?
Olmo 3 32b Think scores higher on coding benchmarks: 38.6 versus 35.2 in the Noometry coding category.
How many benchmarks do Nemotron 4 340b Instruct and Olmo 3 32b Think share?
14 benchmarks have published results for both models. Nemotron 4 340b Instruct has 17 scored results on Noometry and Olmo 3 32b Think has 14.