Model comparison
GLM-5.2 vs GPT-5.4 mini
GLM-5.2 is the stronger model overall, scoring 51.1 to 45.0 on the Noometry Index.
Last verified . 39 shared benchmarks.
Summary
- They share 39 benchmarks with published results for both. GLM-5.2 scores higher in 9 categories and GPT-5.4 mini in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-5.2 leads 42.3 to 30.4.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 62.6% for GLM-5.2 and 37.9% for GPT-5.4 mini.
- GPT-5.4 mini is cheaper at $0.75 / $4.50 per million input/output tokens, against $1.40 / $4.40 for GLM-5.2.
- GLM-5.2 accepts more context: 1M tokens versus 400K.
- GLM-5.2 has downloadable open weights; the other is API-only.
Side by side
| GLM-5.2 | GPT-5.4 mini | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 51.1 | 45.0 |
| Released | 2026-06-13 | 2026-03-17 |
| Weights | Open | Proprietary |
| Context window | 1M | 400K |
| Max output | 131K | 128K |
| Input $ / M tokens | $1.40 | $0.75 |
| Output $ / M tokens | $4.40 | $4.50 |
| Results tracked | 51 | 46 |
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-5.4 mini: 45.2 (#72)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| FrontierCode | 24.5% | 27% |
| LMArena WebDev | 1603 | 1397 |
| SciCode | 50.5% | 49.9% |
| WeirdML | 70.1% | 60.3% |
| LMArena Coding | 1485 | 1438 |
| ALE-Bench | 1,047 | 1,189 |
| SWE-bench Verified | 78.7% | — |
| DeepSWE | 43.8% | — |
Agentic & Tool Use GLM-5.2 leads
GLM-5.2: 32.4 (#63), GPT-5.4 mini: 29.9 (#81)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| APEX-Agents | 45.2% | — |
| τ²-bench Banking | 37.1% | — |
| DeepResearch Bench | — | 36.3% |
| PostTrainBench | 31.7% | — |
| GBAEval | 0% | — |
| Vending-Bench 2 | 8,314 | — |
Reasoning GLM-5.2 leads
GLM-5.2: 42.3 (#52), GPT-5.4 mini: 30.4 (#85)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| ARC-AGI-2 | 22.8% | 18.9% |
| Kagi LLM Benchmark | 62.6% | 37.9% |
| NYT Connections (extended) | 74.3% | 61.8% |
| ARC-AGI-1 | 77% | 63.7% |
| CritPt | 20.9% | 10% |
| Chess Puzzles | 21% | 24% |
| LMArena Hard Prompts | 1480 | 1424 |
| Mystery Game Puzzles | 19% | 11% |
| DTBench | 93.6% | 80% |
| LMCA | 45.8% | 40.8% |
| Epoch Capabilities Index | 151.78 | 148.84 |
| SimpleBench | 58.8% | — |
| Thematic Generalization | — | 61.7% |
| EBR-Bench | 9.5% | — |
| Surface Evolver Bench | 55.6% | — |
| ForecastBench | — | 57 |
Math GLM-5.2 leads
GLM-5.2: 55.7 (#43), GPT-5.4 mini: 45.5 (#75)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.2% | 51.2% |
| FrontierMath Tier 4 | 29.3% | 9.8% |
| OTIS Mock AIME 2024-2025 | 86.4% | 88.9% |
| ProofBench | 35% | 21% |
| LMArena Math | 1482 | 1419 |
| MathArena Final-Answer Competitions | 67.6% | — |
| FrontierMath (Feb 2025 set) | — | 28.3% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge GLM-5.2 leads
GLM-5.2: 57.1 (#40), GPT-5.4 mini: 51.5 (#67)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| GPQA Diamond | 91.9% | 86.9% |
| SimpleQA Verified | 34.2% | 29.4% |
| LMArena Expert | 1486 | 1435 |
| Vectara Hallucination Rate | — | 5.5% |
Multimodal Not comparable
GLM-5.2: —, GPT-5.4 mini: 39.7 (#56)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| LMArena Vision | — | 1245 |
Multilingual GLM-5.2 leads
GLM-5.2: 55.8 (#26), GPT-5.4 mini: 51.9 (#96)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| LMArena Non-English | 1459 | 1405 |
| LMArena Chinese | 1519 | 1446 |
| LMArena French | 1479 | 1440 |
| LMArena German | 1468 | 1409 |
| LMArena Japanese | 1451 | 1374 |
| LMArena Korean | 1445 | 1368 |
| LMArena Russian | 1466 | 1417 |
| LMArena Spanish | 1477 | 1405 |
Instruction Following GLM-5.2 leads
GLM-5.2: 76.9 (#34), GPT-5.4 mini: 74.1 (#102)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| LMArena Instruction Following | 1465 | 1405 |
Long Context GLM-5.2 leads
GLM-5.2: 45.3 (#43), GPT-5.4 mini: 43.0 (#112)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| LMArena Longer Query | 1479 | 1407 |
Writing & Preference GLM-5.2 leads
GLM-5.2: 70.4 (#21), GPT-5.4 mini: 64.0 (#58)
| Benchmark | GLM-5.2 | GPT-5.4 mini |
|---|---|---|
| LMArena Text | 1470 | 1412 |
| LMArena Creative Writing | 1462 | 1370 |
| EQ-Bench Creative Writing | 1757 | 1665 |
| LMArena Multi-Turn | 1469 | 1429 |
| EQ-Bench 4 | 1222 | — |
Frequently asked questions
Is GLM-5.2 better than GPT-5.4 mini?
GLM-5.2 is the stronger model overall, scoring 51.1 to 45.0 on the Noometry Index.
Which is cheaper, GLM-5.2 or GPT-5.4 mini?
GPT-5.4 mini is cheaper. It lists at $0.75 per million input tokens and $4.50 per million output tokens; GLM-5.2 lists at $1.40 and $4.40.
Is GLM-5.2 or GPT-5.4 mini better for coding?
GLM-5.2 scores higher on coding benchmarks: 51.3 versus 45.2 in the Noometry coding category.
Which has the bigger context window?
GLM-5.2 does, with 1M tokens against 400K.
How many benchmarks do GLM-5.2 and GPT-5.4 mini share?
39 benchmarks have published results for both models. GLM-5.2 has 51 scored results on Noometry and GPT-5.4 mini has 46.