Model comparison
Amazon Nova Micro vs gpt-oss-20b
gpt-oss-20b is the stronger model overall, scoring 32.5 to 30.4 on the Noometry Index.
Last verified . 21 shared benchmarks.
Summary
- They share 21 benchmarks with published results for both. Amazon Nova Micro scores higher in 2 categories and gpt-oss-20b in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Amazon Nova Micro leads 22.1 to 9.3.
- The biggest single-benchmark swing is Omni-MATH: 21.4% for Amazon Nova Micro and 56.5% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.035 / $0.14 for Amazon Nova Micro.
- gpt-oss-20b accepts more context: 131K tokens versus 128K.
- gpt-oss-20b has downloadable open weights; the other is API-only.
Side by side
| Amazon Nova Micro | gpt-oss-20b | |
|---|---|---|
| Provider | Amazon | OpenAI |
| Noometry Index | 30.4 | 32.5 |
| Released | 2024-12-03 | 2025-08-05 |
| Weights | Proprietary | Open |
| Context window | 128K | 131K |
| Max output | 10K | 16K |
| Input $ / M tokens | $0.035 | $0.018 |
| Output $ / M tokens | $0.14 | $0.09 |
| Results tracked | 32 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding gpt-oss-20b leads
Amazon Nova Micro: 30.5 (#295), gpt-oss-20b: 37.6 (#192)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| LMArena Coding | 1218 | 1306 |
| SciCode | — | 34.4% |
| WeirdML | — | 40.9% |
| LiveBench Coding | 20.2% | — |
| ALE-Bench | — | 566.05 |
Agentic & Tool Use Amazon Nova Micro leads
Amazon Nova Micro: 22.1 (#132), gpt-oss-20b: 9.3 (#154)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | — | 3.4% |
| Berkeley Function Calling Leaderboard | 22.3% | — |
Reasoning gpt-oss-20b leads
Amazon Nova Micro: 17.4 (#294), gpt-oss-20b: 19.3 (#261)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1274 |
| Kagi LLM Benchmark | — | 53.2% |
| CritPt | — | 1.4% |
| Chess Puzzles | — | 4% |
| LiveBench Reasoning | 25.1% | — |
| DTBench | — | 68% |
| LiveBench Data Analysis | 34% | — |
| LMCA | — | 14.5% |
| Epoch Capabilities Index | — | 137.82 |
| LiveBench | 29.6% | — |
Math gpt-oss-20b leads
Amazon Nova Micro: 26.9 (#254), gpt-oss-20b: 39.4 (#103)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| Omni-MATH | 21.4% | 56.5% |
| LMArena Math | 1206 | 1317 |
| OTIS Mock AIME 2024-2025 | — | 65.3% |
| LiveBench Math | 34.5% | — |
Knowledge gpt-oss-20b leads
Amazon Nova Micro: 29.6 (#237), gpt-oss-20b: 34.6 (#195)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| MMLU-Pro | 51.1% | 74% |
| GPQA (HELM) | 38.3% | 59.4% |
| LMArena Expert | 1184 | 1258 |
| GPQA Diamond | — | 60.8% |
| Vectara Hallucination Rate | 5.5% | — |
| MMLU | 70.8% | — |
Multilingual gpt-oss-20b leads
Amazon Nova Micro: 36.5 (#239), gpt-oss-20b: 42.2 (#197)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | 1186 | 1268 |
| LMArena Chinese | 1209 | 1314 |
| LMArena German | 1192 | 1255 |
| LMArena Japanese | 1154 | 1244 |
| LMArena Korean | 1150 | 1236 |
| LMArena Russian | 1185 | 1278 |
| LMArena Spanish | 1225 | 1267 |
| LMArena French | 1238 | — |
Instruction Following gpt-oss-20b leads
Amazon Nova Micro: 56.3 (#272), gpt-oss-20b: 61.8 (#240)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| IFEval | 76% | 73.2% |
| LMArena Instruction Following | 1174 | 1236 |
| LiveBench Instruction Following | 48% | — |
Long Context gpt-oss-20b leads
Amazon Nova Micro: 36.5 (#229), gpt-oss-20b: 37.9 (#209)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | 1205 | 1250 |
Writing & Preference Amazon Nova Micro leads
Amazon Nova Micro: 39.5 (#247), gpt-oss-20b: 35.5 (#265)
| Benchmark | Amazon Nova Micro | gpt-oss-20b |
|---|---|---|
| LMArena Text | 1208 | 1287 |
| LMArena Creative Writing | 1172 | 1201 |
| WildBench | 74.3% | 73.7% |
| LMArena Multi-Turn | 1178 | 1268 |
| EQ-Bench Creative Writing | — | 666 |
| LiveBench Language | 15.8% | — |
Frequently asked questions
Is Amazon Nova Micro better than gpt-oss-20b?
gpt-oss-20b is the stronger model overall, scoring 32.5 to 30.4 on the Noometry Index.
Which is cheaper, Amazon Nova Micro or gpt-oss-20b?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Amazon Nova Micro lists at $0.035 and $0.14.
Is Amazon Nova Micro or gpt-oss-20b better for coding?
gpt-oss-20b scores higher on coding benchmarks: 37.6 versus 30.5 in the Noometry coding category.
Which has the bigger context window?
gpt-oss-20b does, with 131K tokens against 128K.
How many benchmarks do Amazon Nova Micro and gpt-oss-20b share?
21 benchmarks have published results for both models. Amazon Nova Micro has 32 scored results on Noometry and gpt-oss-20b has 34.