Model comparison
GLM-5 vs GPT-3.5-turbo
GLM-5 is the stronger model overall, scoring 46.1 to 23.2 on the Noometry Index. GPT-3.5-turbo costs 2.1× less per token, which makes it the better buy when GLM-5's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
Summary
- They share 24 benchmarks with published results for both. GLM-5 scores higher in 8 categories and GPT-3.5-turbo in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5 leads 52.3 to 10.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 80% for GLM-5 and 2.2% for GPT-3.5-turbo.
- GPT-3.5-turbo is cheaper at $0.50 / $1.50 per million input/output tokens, against $1 / $3.20 for GLM-5.
- GLM-5 accepts more context: 205K tokens versus 16K.
- GLM-5 has downloadable open weights; the other is API-only.
Side by side
| GLM-5 | GPT-3.5-turbo | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 46.1 | 23.2 |
| Released | 2026-02-11 | 2023-03-01 |
| Weights | Open | Proprietary |
| Context window | 205K | 16K |
| Max output | 131K | 4K |
| Input $ / M tokens | $1 | $0.50 |
| Output $ / M tokens | $3.20 | $1.50 |
| Results tracked | 45 | 44 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5 leads
GLM-5: 49.0 (#52), GPT-3.5-turbo: 23.9 (#331)
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| WeirdML | 48.2% | 3.5% |
| LMArena Coding | 1461 | 1136 |
| SWE-bench Verified | 72.1% | — |
| SWE-bench Verified (bash only) | 72.8% | — |
| LMArena WebDev | 1434 | — |
| SWE-bench Multilingual | 69.7% | — |
| BigCodeBench Instruct | — | 39.1% |
| BigCodeBench Complete | — | 50.6% |
| ALE-Bench | 765.62 | — |
| HumanEval+ | — | 70.7% |
| MBPP+ | — | 69.7% |
Agentic & Tool Use Not comparable
GLM-5: 31.1 (#71), GPT-3.5-turbo: —
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| Terminal-Bench | 52.4% | — |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Banking | 9.8% | — |
| τ²-bench Retail | 73.7% | — |
| τ²-bench Telecom | 86.8% | — |
| METR Time Horizons | — | 21.5% |
| Vending-Bench 2 | 4,432 | — |
Reasoning GLM-5 leads
GLM-5: 27.6 (#116), GPT-3.5-turbo: 13.8 (#332)
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| Chess Puzzles | 10% | 0% |
| LMArena Hard Prompts | 1452 | 1108 |
| Epoch Capabilities Index | 145.83 | 118.55 |
| ForecastBench | 61 | 50.4 |
| ARC-AGI-2 | 4.9% | — |
| SimpleBench | 53.2% | — |
| Kagi LLM Benchmark | 75% | — |
| NYT Connections (extended) | 74.8% | — |
| ARC-AGI-1 | 44.7% | — |
| Mystery Game Puzzles | — | 3% |
| DTBench | — | 48.5% |
| LMCA | — | 9.7% |
| Adversarial NLI | — | 58.1% |
| BIG-Bench Hard | — | 61.6% |
| CommonsenseQA 2.0 | — | 57% |
| WinoGrande | — | 81.6% |
Math GLM-5 leads
GLM-5: 46.4 (#71), GPT-3.5-turbo: 6.3 (#327)
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 80% | 2.2% |
| LMArena Math | 1440 | 1142 |
| FrontierMath (Tiers 1-3) | — | 0% |
| MathArena Final-Answer Competitions | 65.7% | — |
| MATH Level 5 | — | 15.9% |
| FrontierMath (Feb 2025 set) | 16.4% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
| GSM8K | — | 57.8% |
Knowledge GLM-5 leads
GLM-5: 52.3 (#64), GPT-3.5-turbo: 10.0 (#303)
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| GPQA Diamond | 87.8% | 28% |
| LMArena Expert | 1454 | 1070 |
| Vectara Hallucination Rate | 10.1% | — |
| ARC (AI2) Challenge | — | 87.4% |
| BoolQ | — | 87% |
| MMLU | — | 71.4% |
| OpenBookQA | — | 86% |
| TriviaQA | — | 85.8% |
Multilingual GLM-5 leads
GLM-5: 53.7 (#58), GPT-3.5-turbo: 31.5 (#258)
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| LMArena Non-English | 1430 | 1108 |
| LMArena Chinese | 1511 | 1075 |
| LMArena French | 1455 | 1118 |
| LMArena German | 1445 | 1090 |
| LMArena Japanese | 1416 | 1043 |
| LMArena Korean | 1423 | 1019 |
| LMArena Russian | 1436 | 1123 |
| LMArena Spanish | 1454 | 1121 |
Instruction Following GLM-5 leads
GLM-5: 75.2 (#67), GPT-3.5-turbo: 57.9 (#262)
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| LMArena Instruction Following | 1428 | 1119 |
Long Context GLM-5 leads
GLM-5: 44.7 (#60), GPT-3.5-turbo: 34.0 (#254)
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| LMArena Longer Query | 1446 | 1121 |
| CL-bench | 18.7% | — |
Writing & Preference GLM-5 leads
GLM-5: 66.0 (#38), GPT-3.5-turbo: 25.3 (#305)
| Benchmark | GLM-5 | GPT-3.5-turbo |
|---|---|---|
| LMArena Text | 1446 | 1125 |
| LMArena Creative Writing | 1439 | 1092 |
| EQ-Bench Creative Writing | 1601 | 451 |
| LMArena Multi-Turn | 1456 | 1117 |
Frequently asked questions
Is GLM-5 better than GPT-3.5-turbo?
GLM-5 is the stronger model overall, scoring 46.1 to 23.2 on the Noometry Index. GPT-3.5-turbo costs 2.1× less per token, which makes it the better buy when GLM-5's lead doesn't matter for your workload.
Which is cheaper, GLM-5 or GPT-3.5-turbo?
GPT-3.5-turbo is cheaper. It lists at $0.50 per million input tokens and $1.50 per million output tokens; GLM-5 lists at $1 and $3.20.
Is GLM-5 or GPT-3.5-turbo better for coding?
GLM-5 scores higher on coding benchmarks: 49.0 versus 23.9 in the Noometry coding category.
Which has the bigger context window?
GLM-5 does, with 205K tokens against 16K.
How many benchmarks do GLM-5 and GPT-3.5-turbo share?
24 benchmarks have published results for both models. GLM-5 has 45 scored results on Noometry and GPT-3.5-turbo has 44.