Model comparison
GLM-4.5 vs GPT-4o mini
GLM-4.5 is the stronger model overall, scoring 41.4 to 25.4 on the Noometry Index. GPT-4o mini costs 3.8× less per token, which makes it the better buy when GLM-4.5's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
Summary
- They share 22 benchmarks with published results for both. GLM-4.5 scores higher in 7 categories and GPT-4o mini in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-4.5 leads 36.3 to 9.9.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 57.9% for GLM-4.5 and 28.8% for GPT-4o mini.
- GPT-4o mini is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.60 / $2.20 for GLM-4.5.
- GLM-4.5 accepts more context: 131K tokens versus 128K.
- GLM-4.5 has downloadable open weights; the other is API-only.
Side by side
| GLM-4.5 | GPT-4o mini | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 41.4 | 25.4 |
| Released | 2025-07-27 | 2024-07-18 |
| Weights | Open | Proprietary |
| Context window | 131K | 128K |
| Max output | 98K | 16K |
| Input $ / M tokens | $0.60 | $0.15 |
| Output $ / M tokens | $2.20 | $0.60 |
| Results tracked | 27 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-4.5 leads
GLM-4.5: 41.1 (#121), GPT-4o mini: 21.6 (#330)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| WeirdML | 40.6% | 11.8% |
| LMArena Coding | 1434 | 1290 |
| SWE-bench Verified (bash only) | 54.2% | — |
| Aider Polyglot | — | 3.6% |
| BigCodeBench Instruct | — | 46.1% |
| LiveBench Coding | — | 43.1% |
| BigCodeBench Complete | — | 57.4% |
| ALE-Bench | 344.82 | — |
| AlgoTune | 1.52 | — |
| HumanEval+ | — | 83.5% |
| MBPP+ | — | 72.2% |
Agentic & Tool Use Not comparable
GLM-4.5: —, GPT-4o mini: 27.2 (#95)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| BALROG | — | 17.4% |
Reasoning GLM-4.5 leads
GLM-4.5: 27.7 (#99), GPT-4o mini: 8.7 (#343)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| Kagi LLM Benchmark | 57.9% | 28.8% |
| LMArena Hard Prompts | 1430 | 1267 |
| ARC-AGI-2 | — | 0% |
| SimpleBench | — | 10.7% |
| Chess Puzzles | — | 0% |
| LiveBench Reasoning | — | 32.8% |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 54.4% |
| LiveBench Data Analysis | — | 50% |
| LMCA | — | 10.4% |
| Epoch Capabilities Index | — | 126.56 |
| LiveBench | — | 41.3% |
| PIQA | — | 88.7% |
Math GLM-4.5 leads
GLM-4.5: 36.3 (#119), GPT-4o mini: 9.9 (#307)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| LMArena Math | 1427 | 1267 |
| FrontierMath (Tiers 1-3) | — | 0.7% |
| OTIS Mock AIME 2024-2025 | — | 6.9% |
| Omni-MATH | — | 28% |
| LiveBench Math | — | 36.3% |
| MATH Level 5 | — | 52.6% |
| GSM8K | — | 91.3% |
Knowledge GLM-4.5 leads
GLM-4.5: 35.5 (#172), GPT-4o mini: 17.7 (#279)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| Confabulations | 11.3% | 37.2% |
| LMArena Expert | 1433 | 1235 |
| GPQA Diamond | — | 37.7% |
| Humanity's Last Exam | 8.3% | — |
| SimpleQA Verified | — | 8.3% |
| MMLU-Pro | — | 60.3% |
| GPQA (HELM) | — | 36.8% |
| BoolQ | — | 88.7% |
| MMLU | — | 81.8% |
Multimodal Not comparable
GLM-4.5: —, GPT-4o mini: 25.3 (#117)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| LMArena Vision | — | 1066 |
| Video-MME | — | 64.8% |
| GeoBench | — | 64% |
| VPCT | — | 34% |
Multilingual GLM-4.5 leads
GLM-4.5: 53.0 (#74), GPT-4o mini: 42.2 (#194)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| LMArena Non-English | 1417 | 1266 |
| LMArena Chinese | 1465 | 1264 |
| LMArena French | 1423 | 1296 |
| LMArena German | 1406 | 1272 |
| LMArena Japanese | 1416 | 1216 |
| LMArena Korean | 1382 | 1195 |
| LMArena Russian | 1414 | 1275 |
| LMArena Spanish | 1454 | 1276 |
Instruction Following GLM-4.5 leads
GLM-4.5: 74.3 (#100), GPT-4o mini: 62.1 (#234)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| LMArena Instruction Following | 1405 | 1259 |
| LiveBench Instruction Following | — | 56.8% |
| IFEval | — | 78.2% |
Long Context Too close to call
GLM-4.5: 38.2 (#196), GPT-4o mini: 39.1 (#181)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| LMArena Longer Query | 1413 | 1289 |
| Fiction.LiveBench | 58.3% | — |
Writing & Preference GLM-4.5 leads
GLM-4.5: 57.7 (#121), GPT-4o mini: 39.7 (#243)
| Benchmark | GLM-4.5 | GPT-4o mini |
|---|---|---|
| LMArena Text | 1430 | 1286 |
| LMArena Creative Writing | 1395 | 1268 |
| Short-Story Creative Writing | 73.4% | 67.2% |
| EQ-Bench Creative Writing | 1343 | 873 |
| LMArena Multi-Turn | 1415 | 1285 |
| WildBench | — | 79.1% |
| LiveBench Language | — | 28.6% |
Frequently asked questions
Is GLM-4.5 better than GPT-4o mini?
GLM-4.5 is the stronger model overall, scoring 41.4 to 25.4 on the Noometry Index. GPT-4o mini costs 3.8× less per token, which makes it the better buy when GLM-4.5's lead doesn't matter for your workload.
Which is cheaper, GLM-4.5 or GPT-4o mini?
GPT-4o mini is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GLM-4.5 lists at $0.60 and $2.20.
Is GLM-4.5 or GPT-4o mini better for coding?
GLM-4.5 scores higher on coding benchmarks: 41.1 versus 21.6 in the Noometry coding category.
Which has the bigger context window?
GLM-4.5 does, with 131K tokens against 128K.
How many benchmarks do GLM-4.5 and GPT-4o mini share?
22 benchmarks have published results for both models. GLM-4.5 has 27 scored results on Noometry and GPT-4o mini has 60.