Model comparison
GLM-5.2 vs GPT-4.1 nano
GLM-5.2 is the stronger model overall, scoring 50.9 to 27.7 on the Noometry Index. GPT-4.1 nano costs 12× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
Summary
- They share 27 benchmarks with published results for both. GLM-5.2 scores higher in 9 categories and GPT-4.1 nano in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5.2 leads 57.1 to 21.8.
- The biggest single-benchmark swing is ARC-AGI-1: 77% for GLM-5.2 and 0% for GPT-4.1 nano.
- GPT-4.1 nano is cheaper at $0.10 / $0.40 per million input/output tokens, against $1.40 / $4.40 for GLM-5.2.
- GPT-4.1 nano accepts more context: 1.05M tokens versus 1M.
- GLM-5.2 has downloadable open weights; the other is API-only.
Side by side
| GLM-5.2 | GPT-4.1 nano | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 50.9 | 27.7 |
| Released | 2026-06-13 | 2025-04-14 |
| Weights | Open | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 131K | 33K |
| Input $ / M tokens | $1.40 | $0.10 |
| Output $ / M tokens | $4.40 | $0.40 |
| Results tracked | 50 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.2 leads
GLM-5.2: 50.9 (#37), GPT-4.1 nano: 23.8 (#325)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| SciCode | 50.5% | 25.9% |
| WeirdML | 70.1% | 19% |
| LMArena Coding | 1484 | 1306 |
| SWE-bench Verified | 78.7% | — |
| DeepSWE | 43.8% | — |
| FrontierCode | 24.5% | — |
| Aider Polyglot | — | 8.9% |
| LMArena WebDev | 1604 | — |
| ALE-Bench | 1,047 | — |
Agentic & Tool Use GLM-5.2 leads
GLM-5.2: 31.6 (#63), GPT-4.1 nano: 26.3 (#98)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 33% |
| τ²-bench Banking | 37.1% | — |
| PostTrainBench | 31.7% | — |
| GBAEval | 0% | — |
| Vending-Bench 2 | 8,314 | — |
Reasoning GLM-5.2 leads
GLM-5.2: 42.1 (#46), GPT-4.1 nano: 8.6 (#344)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| ARC-AGI-2 | 22.8% | 0% |
| Kagi LLM Benchmark | 62.6% | 33.3% |
| ARC-AGI-1 | 77% | 0% |
| CritPt | 20.9% | 0% |
| LMArena Hard Prompts | 1480 | 1286 |
| DTBench | 93.6% | 52.5% |
| LMCA | 45.8% | 5.5% |
| Epoch Capabilities Index | 151.76 | 129.62 |
| SimpleBench | 58.8% | — |
| NYT Connections (extended) | 74.3% | — |
| Chess Puzzles | 21% | — |
| EBR-Bench | 9.5% | — |
| Mystery Game Puzzles | 19% | — |
| Surface Evolver Bench | 55.6% | — |
Math GLM-5.2 leads
GLM-5.2: 55.4 (#35), GPT-4.1 nano: 25.5 (#246)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.4% | 28.9% |
| LMArena Math | 1482 | 1274 |
| FrontierMath (Tiers 1-3) | 59.2% | — |
| FrontierMath Tier 4 | 29.3% | — |
| MathArena Final-Answer Competitions | 67.6% | — |
| ProofBench | 35% | — |
| Omni-MATH | — | 36.7% |
| MATH Level 5 | — | 70% |
| FrontierMath (Feb 2025 set) | — | 1% |
Knowledge GLM-5.2 leads
GLM-5.2: 57.1 (#36), GPT-4.1 nano: 21.8 (#268)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| GPQA Diamond | 91.9% | 48.9% |
| SimpleQA Verified | 34.2% | 6% |
| LMArena Expert | 1484 | 1272 |
| MMLU-Pro | — | 55% |
| GPQA (HELM) | — | 50.7% |
Multimodal Not comparable
GLM-5.2: —, GPT-4.1 nano: 28.6 (#108)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| LMArena Vision | — | 1063 |
Multilingual GLM-5.2 leads
GLM-5.2: 56.0 (#25), GPT-4.1 nano: 41.8 (#200)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| LMArena Non-English | 1459 | 1260 |
| LMArena Chinese | 1518 | 1270 |
| LMArena German | 1470 | 1288 |
| LMArena Japanese | 1451 | 1199 |
| LMArena Russian | 1467 | 1261 |
| LMArena French | 1476 | — |
| LMArena Korean | 1445 | — |
| LMArena Spanish | 1477 | — |
Instruction Following GLM-5.2 leads
GLM-5.2: 77.1 (#32), GPT-4.1 nano: 68.0 (#188)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| LMArena Instruction Following | 1465 | 1267 |
| IFEval | — | 84.3% |
Long Context GLM-5.2 leads
GLM-5.2: 45.3 (#43), GPT-4.1 nano: 23.7 (#291)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| LMArena Longer Query | 1479 | 1283 |
| Fiction.LiveBench | — | 25% |
Writing & Preference GLM-5.2 leads
GLM-5.2: 70.7 (#20), GPT-4.1 nano: 40.7 (#238)
| Benchmark | GLM-5.2 | GPT-4.1 nano |
|---|---|---|
| LMArena Text | 1470 | 1285 |
| LMArena Creative Writing | 1465 | 1260 |
| EQ-Bench Creative Writing | 1757 | 946 |
| LMArena Multi-Turn | 1468 | 1277 |
| WildBench | — | 81.2% |
| EQ-Bench 4 | 1222 | — |
Frequently asked questions
Is GLM-5.2 better than GPT-4.1 nano?
GLM-5.2 is the stronger model overall, scoring 50.9 to 27.7 on the Noometry Index. GPT-4.1 nano costs 12× 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-4.1 nano?
GPT-4.1 nano is cheaper. It lists at $0.10 per million input tokens and $0.40 per million output tokens; GLM-5.2 lists at $1.40 and $4.40.
Is GLM-5.2 or GPT-4.1 nano better for coding?
GLM-5.2 scores higher on coding benchmarks: 50.9 versus 23.8 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 nano does, with 1.05M tokens against 1M.
How many benchmarks do GLM-5.2 and GPT-4.1 nano share?
27 benchmarks have published results for both models. GLM-5.2 has 50 scored results on Noometry and GPT-4.1 nano has 38.