Model comparison
GLM-5.1 vs GPT-5.4 mini
GLM-5.1 is the stronger model overall, scoring 47.8 to 45.0 on the Noometry Index.
Last verified . 34 shared benchmarks.
Summary
- They share 34 benchmarks with published results for both. GLM-5.1 scores higher in 8 categories and GPT-5.4 mini in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-5.1 leads 39.1 to 30.4.
- The biggest single-benchmark swing is NYT Connections (extended): 77.7% for GLM-5.1 and 61.8% 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-5.1.
- GPT-5.4 mini accepts more context: 400K tokens versus 200K.
- GLM-5.1 has downloadable open weights; the other is API-only.
Side by side
| GLM-5.1 | GPT-5.4 mini | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 47.8 | 45.0 |
| Released | 2026-04-07 | 2026-03-17 |
| Weights | Open | Proprietary |
| Context window | 200K | 400K |
| Max output | 131K | 128K |
| Input $ / M tokens | $1.40 | $0.75 |
| Output $ / M tokens | $4.40 | $4.50 |
| Results tracked | 41 | 46 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.1 leads
GLM-5.1: 48.7 (#55), GPT-5.4 mini: 45.2 (#72)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| LMArena WebDev | 1508 | 1397 |
| SciCode | 43.8% | 49.9% |
| WeirdML | 57.1% | 60.3% |
| LMArena Coding | 1485 | 1438 |
| ALE-Bench | 887.1 | 1,189 |
| SWE-bench Verified | 74.2% | — |
| FrontierCode | — | 27% |
Agentic & Tool Use GPT-5.4 mini leads
GLM-5.1: 24.9 (#113), GPT-5.4 mini: 29.9 (#81)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| APEX-Agents | 40.9% | — |
| DeepResearch Bench | — | 36.3% |
| ExploitBench | 18.1% | — |
| GBAEval | 0% | — |
| Vending-Bench 2 | 5,634 | — |
Reasoning GLM-5.1 leads
GLM-5.1: 39.1 (#60), GPT-5.4 mini: 30.4 (#85)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| NYT Connections (extended) | 77.7% | 61.8% |
| CritPt | 4.6% | 10% |
| Chess Puzzles | 19% | 24% |
| Thematic Generalization | 69.8% | 61.7% |
| LMArena Hard Prompts | 1472 | 1424 |
| Epoch Capabilities Index | 149.84 | 148.84 |
| ARC-AGI-2 | — | 18.9% |
| SimpleBench | 55.1% | — |
| Kagi LLM Benchmark | — | 37.9% |
| ARC-AGI-1 | — | 63.7% |
| Mystery Game Puzzles | — | 11% |
| DTBench | — | 80% |
| LMCA | — | 40.8% |
| ForecastBench | — | 57 |
Math GLM-5.1 leads
GLM-5.1: 49.7 (#60), GPT-5.4 mini: 45.5 (#75)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 36.8% | 51.2% |
| OTIS Mock AIME 2024-2025 | 93.3% | 88.9% |
| ProofBench | 22.2% | 21% |
| LMArena Math | 1473 | 1419 |
| FrontierMath (Feb 2025 set) | 33.4% | 28.3% |
| FrontierMath Tier 4 (v1) | 12.5% | 2.1% |
| FrontierMath Tier 4 | — | 9.8% |
| MathArena Final-Answer Competitions | 67.1% | — |
Knowledge GLM-5.1 leads
GLM-5.1: 54.9 (#50), GPT-5.4 mini: 51.5 (#67)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| GPQA Diamond | 89.9% | 86.9% |
| SimpleQA Verified | 34% | 29.4% |
| LMArena Expert | 1476 | 1435 |
| Vectara Hallucination Rate | — | 5.5% |
Multimodal Not comparable
GLM-5.1: —, GPT-5.4 mini: 39.7 (#56)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| LMArena Vision | — | 1245 |
Multilingual GLM-5.1 leads
GLM-5.1: 55.0 (#36), GPT-5.4 mini: 51.9 (#96)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| LMArena Non-English | 1447 | 1405 |
| LMArena Chinese | 1515 | 1446 |
| LMArena French | 1474 | 1440 |
| LMArena German | 1465 | 1409 |
| LMArena Japanese | 1434 | 1374 |
| LMArena Korean | 1418 | 1368 |
| LMArena Russian | 1454 | 1417 |
| LMArena Spanish | 1469 | 1405 |
Instruction Following GLM-5.1 leads
GLM-5.1: 76.3 (#42), GPT-5.4 mini: 74.1 (#102)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| LMArena Instruction Following | 1451 | 1405 |
Long Context GLM-5.1 leads
GLM-5.1: 44.9 (#53), GPT-5.4 mini: 43.0 (#112)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| LMArena Longer Query | 1466 | 1407 |
Writing & Preference GLM-5.1 leads
GLM-5.1: 66.9 (#31), GPT-5.4 mini: 64.0 (#58)
| Benchmark | GLM-5.1 | GPT-5.4 mini |
|---|---|---|
| LMArena Text | 1461 | 1412 |
| LMArena Creative Writing | 1453 | 1370 |
| EQ-Bench Creative Writing | 1592 | 1665 |
| LMArena Multi-Turn | 1472 | 1429 |
Frequently asked questions
Is GLM-5.1 better than GPT-5.4 mini?
GLM-5.1 is the stronger model overall, scoring 47.8 to 45.0 on the Noometry Index.
Which is cheaper, GLM-5.1 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-5.1 lists at $1.40 and $4.40.
Is GLM-5.1 or GPT-5.4 mini better for coding?
GLM-5.1 scores higher on coding benchmarks: 48.7 versus 45.2 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 200K.
How many benchmarks do GLM-5.1 and GPT-5.4 mini share?
34 benchmarks have published results for both models. GLM-5.1 has 41 scored results on Noometry and GPT-5.4 mini has 46.