Model comparison
GLM-4.6 vs GPT-5
GPT-5 is the stronger model overall, scoring 50.9 to 41.4 on the Noometry Index. GLM-4.6 costs 3.4× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Last verified . 28 shared benchmarks.
Summary
- They share 28 benchmarks with published results for both. GLM-4.6 scores higher in 2 categories and GPT-5 in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 43.4.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 47.4% for GLM-4.6 and 72.7% for GPT-5.
- GLM-4.6 is cheaper at $0.60 / $2.20 per million input/output tokens, against $1.25 / $10 for GPT-5.
- GPT-5 accepts more context: 400K tokens versus 205K.
- GLM-4.6 has downloadable open weights; the other is API-only.
Side by side
| GLM-4.6 | GPT-5 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 41.4 | 50.9 |
| Released | 2025-09-30 | 2025-08-07 |
| Weights | Open | Proprietary |
| Context window | 205K | 400K |
| Max output | 131K | 128K |
| Input $ / M tokens | $0.60 | $1.25 |
| Output $ / M tokens | $2.20 | $10 |
| Results tracked | 29 | 69 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 leads
GLM-4.6: 40.1 (#148), GPT-5: 50.3 (#47)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| SWE-bench Verified (bash only) | 55.4% | 65% |
| LMArena WebDev | 1340 | 1418 |
| SciCode | 38.4% | 42.9% |
| LMArena Coding | 1449 | 1436 |
| ALE-Bench | 340.82 | 1,162 |
| SWE-bench Verified | — | 73.6% |
| Aider Polyglot | — | 88% |
| GSO | — | 6.9% |
| WeirdML | — | 60.7% |
| AlgoTune | — | 1.67 |
Agentic & Tool Use Too close to call
GLM-4.6: 32.3 (#66), GPT-5: 33.1 (#56)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| Terminal-Bench | 24.5% | 49.6% |
| Berkeley Function Calling Leaderboard | 72.4% | — |
| GDPval | — | 34.8% |
| Remote Labor Index | — | 1.7% |
| DeepResearch Bench | — | 49.6% |
| BALROG | — | 32.8% |
| LMArena Search | — | 1133 |
| METR Time Horizons | — | 69.6% |
Reasoning GPT-5 leads
GLM-4.6: 23.7 (#172), GPT-5: 38.3 (#64)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| Kagi LLM Benchmark | 47.4% | 72.7% |
| CritPt | 1.1% | 12.6% |
| LMArena Hard Prompts | 1440 | 1416 |
| ARC-AGI-2 | — | 9.9% |
| SimpleBench | — | 56.7% |
| ARC-AGI-1 | — | 65.7% |
| Chess Puzzles | — | 37% |
| EnigmaEval | — | 10.5% |
| EBR-Bench | — | 12.7% |
| Mystery Game Puzzles | — | 23% |
| DTBench | — | 90.7% |
| LMCA | — | 40% |
| Epoch Capabilities Index | — | 150 |
| ForecastBench | — | 61.4 |
Math GPT-5 leads
GLM-4.6: 39.1 (#111), GPT-5: 55.0 (#44)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| LMArena Math | 1432 | 1407 |
| FrontierMath (Feb 2025 set) | 3.8% | 32.4% |
| FrontierMath Tier 4 (v1) | 2.1% | 12.5% |
| FrontierMath (Tiers 1-3) | — | 55.4% |
| FrontierMath Tier 4 | — | 22% |
| OTIS Mock AIME 2024-2025 | — | 91.4% |
| ProofBench | — | 18% |
| Omni-MATH | — | 64.7% |
| MATH Level 5 | — | 98.1% |
Knowledge GPT-5 leads
GLM-4.6: 40.2 (#124), GPT-5: 56.6 (#43)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| Vectara Hallucination Rate | 9.5% | 14.7% |
| LMArena Expert | 1431 | 1419 |
| GPQA Diamond | — | 86.2% |
| Humanity's Last Exam | — | 25.3% |
| SimpleQA Verified | — | 50.1% |
| MMLU-Pro | — | 86.3% |
| Confabulations | — | 10.3% |
| GPQA (HELM) | — | 79.2% |
Multimodal Not comparable
GLM-4.6: —, GPT-5: 46.8 (#13)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| LMArena Vision | — | 1232 |
| GeoBench | — | 81% |
| VPCT | — | 66% |
Multilingual GLM-4.6 leads
GLM-4.6: 53.5 (#66), GPT-5: 51.4 (#110)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| LMArena Non-English | 1426 | 1397 |
| LMArena Chinese | 1499 | 1422 |
| LMArena French | 1459 | 1410 |
| LMArena German | 1447 | 1416 |
| LMArena Japanese | 1393 | 1409 |
| LMArena Korean | 1400 | 1360 |
| LMArena Russian | 1419 | 1406 |
| LMArena Spanish | 1436 | 1399 |
Instruction Following Too close to call
GLM-4.6: 74.3 (#98), GPT-5: 73.8 (#113)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| LMArena Instruction Following | 1410 | 1388 |
| IFEval | — | 87.5% |
Long Context GPT-5 leads
GLM-4.6: 43.4 (#94), GPT-5: 69.5 (#2)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| LMArena Longer Query | 1422 | 1399 |
| Fiction.LiveBench | — | 97.2% |
Writing & Preference GPT-5 leads
GLM-4.6: 61.1 (#90), GPT-5: 63.4 (#65)
| Benchmark | GLM-4.6 | GPT-5 |
|---|---|---|
| LMArena Text | 1440 | 1406 |
| LMArena Creative Writing | 1411 | 1365 |
| EQ-Bench Creative Writing | 1411 | 1627 |
| LMArena Multi-Turn | 1427 | 1426 |
| Short-Story Creative Writing | — | 86% |
| WildBench | — | 85.7% |
Frequently asked questions
Is GLM-4.6 better than GPT-5?
GPT-5 is the stronger model overall, scoring 50.9 to 41.4 on the Noometry Index. GLM-4.6 costs 3.4× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Which is cheaper, GLM-4.6 or GPT-5?
GLM-4.6 is cheaper. It lists at $0.60 per million input tokens and $2.20 per million output tokens; GPT-5 lists at $1.25 and $10.
Is GLM-4.6 or GPT-5 better for coding?
GPT-5 scores higher on coding benchmarks: 50.3 versus 40.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5 does, with 400K tokens against 205K.
How many benchmarks do GLM-4.6 and GPT-5 share?
28 benchmarks have published results for both models. GLM-4.6 has 29 scored results on Noometry and GPT-5 has 69.