Model comparison
gpt-oss-20b vs Grok 4.5
Grok 4.5 is the stronger model overall, scoring 55.0 to 32.5 on the Noometry Index. gpt-oss-20b costs 83× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Last verified . 28 shared benchmarks.
Summary
- They share 28 benchmarks with published results for both. gpt-oss-20b scores higher in 0 categories and Grok 4.5 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.5 leads 56.1 to 19.3.
- The biggest single-benchmark swing is GPQA Diamond: 60.8% for gpt-oss-20b and 93.4% for Grok 4.5.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $2 / $6 for Grok 4.5.
- Grok 4.5 accepts more context: 500K tokens versus 131K.
- gpt-oss-20b has downloadable open weights; the other is API-only.
Side by side
| gpt-oss-20b | Grok 4.5 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 32.5 | 55.0 |
| Released | 2025-08-05 | 2026-07-08 |
| Weights | Open | Proprietary |
| Context window | 131K | 500K |
| Max output | 16K | 500K |
| Input $ / M tokens | $0.018 | $2 |
| Output $ / M tokens | $0.09 | $6 |
| Results tracked | 34 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.5 leads
gpt-oss-20b: 37.6 (#192), Grok 4.5: 52.2 (#35)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| SciCode | 34.4% | 54.1% |
| WeirdML | 40.9% | 46.4% |
| LMArena Coding | 1306 | 1474 |
| ALE-Bench | 566.05 | 1,309 |
| DeepSWE | — | 53.8% |
| FrontierCode | — | 42.4% |
| LMArena WebDev | — | 1553 |
Agentic & Tool Use Grok 4.5 leads
gpt-oss-20b: 9.3 (#154), Grok 4.5: 44.4 (#17)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| Terminal-Bench | 3.4% | — |
| APEX-Agents | — | 56.2% |
| τ²-bench Banking | — | 47.9% |
| PostTrainBench | — | 23.4% |
| GBAEval | — | 65.4% |
| GDP.pdf | — | 14% |
| LMArena Search | — | 1213 |
| Vending-Bench 2 | — | 3,887 |
Reasoning Grok 4.5 leads
gpt-oss-20b: 19.3 (#261), Grok 4.5: 56.1 (#25)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| Kagi LLM Benchmark | 53.2% | 83.5% |
| CritPt | 1.4% | 15.4% |
| Chess Puzzles | 4% | 36% |
| LMArena Hard Prompts | 1274 | 1462 |
| DTBench | 68% | 96.5% |
| LMCA | 14.5% | 45.2% |
| Epoch Capabilities Index | 137.82 | 153.92 |
| ARC-AGI-2 | — | 52.6% |
| SimpleBench | — | 70% |
| NYT Connections (extended) | — | 79.9% |
| ARC-AGI-1 | — | 87.2% |
| Surface Evolver Bench | — | 74.4% |
Math Grok 4.5 leads
gpt-oss-20b: 39.4 (#103), Grok 4.5: 60.9 (#35)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | 97.8% |
| LMArena Math | 1317 | 1459 |
| FrontierMath (Tiers 1-3) | — | 57.2% |
| FrontierMath Tier 4 | — | 24.4% |
| ProofBench | — | 31% |
| Omni-MATH | 56.5% | — |
Knowledge Grok 4.5 leads
gpt-oss-20b: 34.6 (#195), Grok 4.5: 62.3 (#24)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 60.8% | 93.4% |
| LMArena Expert | 1258 | 1466 |
| SimpleQA Verified | — | 48.3% |
| MMLU-Pro | 74% | — |
| GPQA (HELM) | 59.4% | — |
Multimodal Not comparable
gpt-oss-20b: —, Grok 4.5: 37.6 (#72)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| LMArena Vision | — | 1288 |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
| LMArena Document | — | 1452 |
Multilingual Grok 4.5 leads
gpt-oss-20b: 42.2 (#197), Grok 4.5: 54.4 (#42)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1268 | 1440 |
| LMArena Chinese | 1314 | 1496 |
| LMArena German | 1255 | 1446 |
| LMArena Japanese | 1244 | 1428 |
| LMArena Korean | 1236 | 1404 |
| LMArena Russian | 1278 | 1448 |
| LMArena Spanish | 1267 | 1450 |
| LMArena French | — | 1456 |
Instruction Following Grok 4.5 leads
gpt-oss-20b: 61.8 (#240), Grok 4.5: 76.0 (#48)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1236 | 1446 |
| IFEval | 73.2% | — |
Long Context Grok 4.5 leads
gpt-oss-20b: 37.9 (#209), Grok 4.5: 44.8 (#56)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1250 | 1463 |
Writing & Preference Grok 4.5 leads
gpt-oss-20b: 35.5 (#265), Grok 4.5: 65.8 (#42)
| Benchmark | gpt-oss-20b | Grok 4.5 |
|---|---|---|
| LMArena Text | 1287 | 1448 |
| LMArena Creative Writing | 1201 | 1442 |
| EQ-Bench Creative Writing | 666 | 1579 |
| LMArena Multi-Turn | 1268 | 1456 |
| WildBench | 73.7% | — |
Frequently asked questions
Is gpt-oss-20b better than Grok 4.5?
Grok 4.5 is the stronger model overall, scoring 55.0 to 32.5 on the Noometry Index. gpt-oss-20b costs 83× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Which is cheaper, gpt-oss-20b or Grok 4.5?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Grok 4.5 lists at $2 and $6.
Is gpt-oss-20b or Grok 4.5 better for coding?
Grok 4.5 scores higher on coding benchmarks: 52.2 versus 37.6 in the Noometry coding category.
Which has the bigger context window?
Grok 4.5 does, with 500K tokens against 131K.
How many benchmarks do gpt-oss-20b and Grok 4.5 share?
28 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Grok 4.5 has 52.