Model comparison
GPT-4.1 mini vs gpt-oss-20b
GPT-4.1 mini is the stronger model overall, scoring 33.6 to 32.5 on the Noometry Index. gpt-oss-20b costs 19× less per token, which makes it the better buy when GPT-4.1 mini's lead doesn't matter for your workload.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. GPT-4.1 mini scores higher in 5 categories and gpt-oss-20b in 4 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where GPT-4.1 mini leads 33.3 to 9.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 44.7% for GPT-4.1 mini and 65.3% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.40 / $1.60 for GPT-4.1 mini.
- GPT-4.1 mini accepts more context: 1.05M tokens versus 131K.
- gpt-oss-20b has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 mini | gpt-oss-20b | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 33.6 | 32.5 |
| Released | 2025-04-14 | 2025-08-05 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 131K |
| Max output | 33K | 16K |
| Input $ / M tokens | $0.40 | $0.018 |
| Output $ / M tokens | $1.60 | $0.09 |
| Results tracked | 47 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding gpt-oss-20b leads
GPT-4.1 mini: 30.6 (#293), gpt-oss-20b: 37.6 (#192)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| SciCode | 40.4% | 34.4% |
| WeirdML | 37.6% | 40.9% |
| LMArena Coding | 1367 | 1306 |
| SWE-bench Verified (bash only) | 23.9% | — |
| Aider Polyglot | 32.4% | — |
| BigCodeBench Instruct | 48.9% | — |
| CadEval | 16% | — |
| ALE-Bench | — | 566.05 |
Agentic & Tool Use GPT-4.1 mini leads
GPT-4.1 mini: 33.3 (#55), gpt-oss-20b: 9.3 (#154)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | — | 3.4% |
| Berkeley Function Calling Leaderboard | 50.5% | — |
Reasoning gpt-oss-20b leads
GPT-4.1 mini: 10.8 (#340), gpt-oss-20b: 19.3 (#261)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| Kagi LLM Benchmark | 48.6% | 53.2% |
| CritPt | 0% | 1.4% |
| Chess Puzzles | 7% | 4% |
| LMArena Hard Prompts | 1349 | 1274 |
| DTBench | 68.8% | 68% |
| LMCA | 21.1% | 14.5% |
| Epoch Capabilities Index | 135.01 | 137.82 |
| ARC-AGI-2 | 0% | — |
| ARC-AGI-1 | 3.5% | — |
| Mystery Game Puzzles | 7% | — |
Math gpt-oss-20b leads
GPT-4.1 mini: 24.1 (#270), gpt-oss-20b: 39.4 (#103)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 44.7% | 65.3% |
| Omni-MATH | 49.1% | 56.5% |
| LMArena Math | 1343 | 1317 |
| FrontierMath (Tiers 1-3) | 6.7% | — |
| MATH Level 5 | 87.3% | — |
| FrontierMath (Feb 2025 set) | 4.5% | — |
Knowledge Too close to call
GPT-4.1 mini: 34.7 (#194), gpt-oss-20b: 34.6 (#195)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | 65.8% | 60.8% |
| MMLU-Pro | 78.3% | 74% |
| GPQA (HELM) | 61.4% | 59.4% |
| LMArena Expert | 1338 | 1258 |
| SimpleQA Verified | 12.7% | — |
Multimodal Not comparable
GPT-4.1 mini: 35.8 (#82), gpt-oss-20b: —
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| LMArena Vision | 1181 | — |
Multilingual GPT-4.1 mini leads
GPT-4.1 mini: 45.7 (#166), gpt-oss-20b: 42.2 (#197)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | 1318 | 1268 |
| LMArena Chinese | 1329 | 1314 |
| LMArena German | 1351 | 1255 |
| LMArena Japanese | 1290 | 1244 |
| LMArena Korean | 1298 | 1236 |
| LMArena Russian | 1324 | 1278 |
| LMArena Spanish | 1319 | 1267 |
| LMArena French | 1358 | — |
Instruction Following GPT-4.1 mini leads
GPT-4.1 mini: 73.7 (#118), gpt-oss-20b: 61.8 (#240)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| IFEval | 90.4% | 73.2% |
| LMArena Instruction Following | 1333 | 1236 |
Long Context gpt-oss-20b leads
GPT-4.1 mini: 31.8 (#275), gpt-oss-20b: 37.9 (#209)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | 1344 | 1250 |
| Fiction.LiveBench | 44.4% | — |
Writing & Preference GPT-4.1 mini leads
GPT-4.1 mini: 48.6 (#199), gpt-oss-20b: 35.5 (#265)
| Benchmark | GPT-4.1 mini | gpt-oss-20b |
|---|---|---|
| LMArena Text | 1340 | 1287 |
| LMArena Creative Writing | 1300 | 1201 |
| EQ-Bench Creative Writing | 1147 | 666 |
| WildBench | 83.8% | 73.7% |
| LMArena Multi-Turn | 1354 | 1268 |
Frequently asked questions
Is GPT-4.1 mini better than gpt-oss-20b?
GPT-4.1 mini is the stronger model overall, scoring 33.6 to 32.5 on the Noometry Index. gpt-oss-20b costs 19× less per token, which makes it the better buy when GPT-4.1 mini's lead doesn't matter for your workload.
Which is cheaper, GPT-4.1 mini 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; GPT-4.1 mini lists at $0.40 and $1.60.
Is GPT-4.1 mini or gpt-oss-20b better for coding?
gpt-oss-20b scores higher on coding benchmarks: 37.6 versus 30.6 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 mini does, with 1.05M tokens against 131K.
How many benchmarks do GPT-4.1 mini and gpt-oss-20b share?
32 benchmarks have published results for both models. GPT-4.1 mini has 47 scored results on Noometry and gpt-oss-20b has 34.