Model comparison
GLM-4.5 vs GPT-5.4 mini
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 42.0 on the Noometry Index. GLM-4.5 costs 1.7× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Last verified . 21 shared benchmarks.
Summary
- They share 21 benchmarks with published results for both. GLM-4.5 scores higher in 1 category and GPT-5.4 mini in 7 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.4 mini leads 51.5 to 35.9.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 57.9% for GLM-4.5 and 37.9% for GPT-5.4 mini.
- GLM-4.5 is cheaper at $0.60 / $2.20 per million input/output tokens, against $0.75 / $4.50 for GPT-5.4 mini.
- GPT-5.4 mini accepts more context: 400K tokens versus 131K.
- GLM-4.5 has downloadable open weights; the other is API-only.
Side by side
| GLM-4.5 | GPT-5.4 mini | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 42.0 | 45.0 |
| Released | 2025-07-27 | 2026-03-17 |
| Weights | Open | Proprietary |
| Context window | 131K | 400K |
| Max output | 98K | 128K |
| Input $ / M tokens | $0.60 | $0.75 |
| Output $ / M tokens | $2.20 | $4.50 |
| Results tracked | 27 | 46 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 mini leads
GLM-4.5: 41.4 (#125), GPT-5.4 mini: 45.2 (#72)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| WeirdML | 40.6% | 60.3% |
| LMArena Coding | 1434 | 1438 |
| ALE-Bench | 344.82 | 1,189 |
| FrontierCode | — | 27% |
| SWE-bench Verified (bash only) | 54.2% | — |
| LMArena WebDev | — | 1397 |
| SciCode | — | 49.9% |
| AlgoTune | 1.52 | — |
Agentic & Tool Use Not comparable
GLM-4.5: —, GPT-5.4 mini: 29.9 (#81)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| DeepResearch Bench | — | 36.3% |
Reasoning GPT-5.4 mini leads
GLM-4.5: 28.6 (#100), GPT-5.4 mini: 30.4 (#85)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| Kagi LLM Benchmark | 57.9% | 37.9% |
| LMArena Hard Prompts | 1429 | 1424 |
| ARC-AGI-2 | — | 18.9% |
| NYT Connections (extended) | — | 61.8% |
| ARC-AGI-1 | — | 63.7% |
| CritPt | — | 10% |
| Chess Puzzles | — | 24% |
| Thematic Generalization | — | 61.7% |
| Mystery Game Puzzles | — | 11% |
| DTBench | — | 80% |
| LMCA | — | 40.8% |
| Epoch Capabilities Index | — | 148.84 |
| ForecastBench | — | 57 |
Math GPT-5.4 mini leads
GLM-4.5: 39.0 (#116), GPT-5.4 mini: 45.5 (#75)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Math | 1427 | 1419 |
| FrontierMath (Tiers 1-3) | — | 51.2% |
| FrontierMath Tier 4 | — | 9.8% |
| OTIS Mock AIME 2024-2025 | — | 88.9% |
| ProofBench | — | 21% |
| FrontierMath (Feb 2025 set) | — | 28.3% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge GPT-5.4 mini leads
GLM-4.5: 35.9 (#179), GPT-5.4 mini: 51.5 (#67)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Expert | 1433 | 1435 |
| GPQA Diamond | — | 86.9% |
| Humanity's Last Exam | 8.3% | — |
| SimpleQA Verified | — | 29.4% |
| Confabulations | 11.3% | — |
| Vectara Hallucination Rate | — | 5.5% |
Multimodal Not comparable
GLM-4.5: —, GPT-5.4 mini: 39.7 (#56)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Vision | — | 1245 |
Multilingual Too close to call
GLM-4.5: 52.8 (#77), GPT-5.4 mini: 51.9 (#96)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Non-English | 1417 | 1405 |
| LMArena Chinese | 1465 | 1446 |
| LMArena French | 1418 | 1440 |
| LMArena German | 1407 | 1409 |
| LMArena Japanese | 1415 | 1374 |
| LMArena Korean | 1380 | 1368 |
| LMArena Russian | 1414 | 1417 |
| LMArena Spanish | 1454 | 1405 |
Instruction Following Too close to call
GLM-4.5: 74.1 (#104), GPT-5.4 mini: 74.1 (#102)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Instruction Following | 1404 | 1405 |
Long Context GPT-5.4 mini leads
GLM-4.5: 38.2 (#201), GPT-5.4 mini: 43.0 (#112)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Longer Query | 1412 | 1407 |
| Fiction.LiveBench | 58.3% | — |
Writing & Preference GPT-5.4 mini leads
GLM-4.5: 57.5 (#127), GPT-5.4 mini: 64.0 (#58)
| Benchmark | GLM-4.5 | GPT-5.4 mini |
|---|---|---|
| LMArena Text | 1430 | 1412 |
| LMArena Creative Writing | 1395 | 1370 |
| EQ-Bench Creative Writing | 1343 | 1665 |
| LMArena Multi-Turn | 1415 | 1429 |
| Short-Story Creative Writing | 73.4% | — |
Frequently asked questions
Is GLM-4.5 better than GPT-5.4 mini?
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 42.0 on the Noometry Index. GLM-4.5 costs 1.7× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Which is cheaper, GLM-4.5 or GPT-5.4 mini?
GLM-4.5 is cheaper. It lists at $0.60 per million input tokens and $2.20 per million output tokens; GPT-5.4 mini lists at $0.75 and $4.50.
Is GLM-4.5 or GPT-5.4 mini better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.2 versus 41.4 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 131K.
How many benchmarks do GLM-4.5 and GPT-5.4 mini share?
21 benchmarks have published results for both models. GLM-4.5 has 27 scored results on Noometry and GPT-5.4 mini has 46.