Model comparison
GLM-5.2 vs gpt-oss-20b
GLM-5.2 is the stronger model overall, scoring 51.1 to 32.5 on the Noometry Index. gpt-oss-20b costs 60× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.
Last verified . 28 shared benchmarks.
Summary
- They share 28 benchmarks with published results for both. GLM-5.2 scores higher in 9 categories and gpt-oss-20b in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GLM-5.2 leads 70.4 to 35.5.
- The biggest single-benchmark swing is LMCA: 45.8% for GLM-5.2 and 14.5% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $1.40 / $4.40 for GLM-5.2.
- GLM-5.2 accepts more context: 1M tokens versus 131K.
Side by side
| GLM-5.2 | gpt-oss-20b | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 51.1 | 32.5 |
| Released | 2026-06-13 | 2025-08-05 |
| Weights | Open | Open |
| Context window | 1M | 131K |
| Max output | 131K | 16K |
| Input $ / M tokens | $1.40 | $0.018 |
| Output $ / M tokens | $4.40 | $0.09 |
| Results tracked | 51 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.2 leads
GLM-5.2: 51.3 (#41), gpt-oss-20b: 37.6 (#192)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| SciCode | 50.5% | 34.4% |
| WeirdML | 70.1% | 40.9% |
| LMArena Coding | 1485 | 1306 |
| ALE-Bench | 1,047 | 566.05 |
| SWE-bench Verified | 78.7% | — |
| DeepSWE | 43.8% | — |
| FrontierCode | 24.5% | — |
| LMArena WebDev | 1603 | — |
Agentic & Tool Use GLM-5.2 leads
GLM-5.2: 32.4 (#63), gpt-oss-20b: 9.3 (#154)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | — | 3.4% |
| APEX-Agents | 45.2% | — |
| τ²-bench Banking | 37.1% | — |
| PostTrainBench | 31.7% | — |
| GBAEval | 0% | — |
| Vending-Bench 2 | 8,314 | — |
Reasoning GLM-5.2 leads
GLM-5.2: 42.3 (#52), gpt-oss-20b: 19.3 (#261)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| Kagi LLM Benchmark | 62.6% | 53.2% |
| CritPt | 20.9% | 1.4% |
| Chess Puzzles | 21% | 4% |
| LMArena Hard Prompts | 1480 | 1274 |
| DTBench | 93.6% | 68% |
| LMCA | 45.8% | 14.5% |
| Epoch Capabilities Index | 151.78 | 137.82 |
| ARC-AGI-2 | 22.8% | — |
| SimpleBench | 58.8% | — |
| NYT Connections (extended) | 74.3% | — |
| ARC-AGI-1 | 77% | — |
| EBR-Bench | 9.5% | — |
| Mystery Game Puzzles | 19% | — |
| Surface Evolver Bench | 55.6% | — |
Math GLM-5.2 leads
GLM-5.2: 55.7 (#43), gpt-oss-20b: 39.4 (#103)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.4% | 65.3% |
| LMArena Math | 1482 | 1317 |
| FrontierMath (Tiers 1-3) | 59.2% | — |
| FrontierMath Tier 4 | 29.3% | — |
| MathArena Final-Answer Competitions | 67.6% | — |
| ProofBench | 35% | — |
| Omni-MATH | — | 56.5% |
Knowledge GLM-5.2 leads
GLM-5.2: 57.1 (#40), gpt-oss-20b: 34.6 (#195)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | 91.9% | 60.8% |
| LMArena Expert | 1486 | 1258 |
| SimpleQA Verified | 34.2% | — |
| MMLU-Pro | — | 74% |
| GPQA (HELM) | — | 59.4% |
Multilingual GLM-5.2 leads
GLM-5.2: 55.8 (#26), gpt-oss-20b: 42.2 (#197)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | 1459 | 1268 |
| LMArena Chinese | 1519 | 1314 |
| LMArena German | 1468 | 1255 |
| LMArena Japanese | 1451 | 1244 |
| LMArena Korean | 1445 | 1236 |
| LMArena Russian | 1466 | 1278 |
| LMArena Spanish | 1477 | 1267 |
| LMArena French | 1479 | — |
Instruction Following GLM-5.2 leads
GLM-5.2: 76.9 (#34), gpt-oss-20b: 61.8 (#240)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| LMArena Instruction Following | 1465 | 1236 |
| IFEval | — | 73.2% |
Long Context GLM-5.2 leads
GLM-5.2: 45.3 (#43), gpt-oss-20b: 37.9 (#209)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | 1479 | 1250 |
Writing & Preference GLM-5.2 leads
GLM-5.2: 70.4 (#21), gpt-oss-20b: 35.5 (#265)
| Benchmark | GLM-5.2 | gpt-oss-20b |
|---|---|---|
| LMArena Text | 1470 | 1287 |
| LMArena Creative Writing | 1462 | 1201 |
| EQ-Bench Creative Writing | 1757 | 666 |
| LMArena Multi-Turn | 1469 | 1268 |
| WildBench | — | 73.7% |
| EQ-Bench 4 | 1222 | — |
Frequently asked questions
Is GLM-5.2 better than gpt-oss-20b?
GLM-5.2 is the stronger model overall, scoring 51.1 to 32.5 on the Noometry Index. gpt-oss-20b costs 60× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.
Which is cheaper, GLM-5.2 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; GLM-5.2 lists at $1.40 and $4.40.
Is GLM-5.2 or gpt-oss-20b better for coding?
GLM-5.2 scores higher on coding benchmarks: 51.3 versus 37.6 in the Noometry coding category.
Which has the bigger context window?
GLM-5.2 does, with 1M tokens against 131K.
How many benchmarks do GLM-5.2 and gpt-oss-20b share?
28 benchmarks have published results for both models. GLM-5.2 has 51 scored results on Noometry and gpt-oss-20b has 34.