Model comparison
GLM-5.1 vs GPT-5.1
GPT-5.1 is the stronger model overall, scoring 49.0 to 47.8 on the Noometry Index. GLM-5.1 costs 1.6× less per token, which makes it the better buy when GPT-5.1's lead doesn't matter for your workload.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. GLM-5.1 scores higher in 4 categories and GPT-5.1 in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where GPT-5.1 leads 32.7 to 24.9.
- The biggest single-benchmark swing is SimpleQA Verified: 34% for GLM-5.1 and 48% for GPT-5.1.
- GLM-5.1 is cheaper at $1.40 / $4.40 per million input/output tokens, against $1.25 / $10 for GPT-5.1.
- GPT-5.1 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-5.1 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 47.8 | 49.0 |
| Released | 2026-04-07 | 2025-11-13 |
| Weights | Open | Proprietary |
| Context window | 200K | 400K |
| Max output | 131K | 128K |
| Input $ / M tokens | $1.40 | $1.25 |
| Output $ / M tokens | $4.40 | $10 |
| Results tracked | 41 | 63 |
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-5.1: 46.4 (#66)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| SWE-bench Verified | 74.2% | 68% |
| LMArena WebDev | 1508 | 1395 |
| SciCode | 43.8% | 43.3% |
| WeirdML | 57.1% | 60.8% |
| LMArena Coding | 1485 | 1454 |
| ALE-Bench | 887.1 | 1,192 |
| SWE-bench Verified (bash only) | — | 66% |
| GSO | — | 13.7% |
| LiveBench Coding | — | 72.5% |
Agentic & Tool Use GPT-5.1 leads
GLM-5.1: 24.9 (#113), GPT-5.1: 32.7 (#60)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| Vending-Bench 2 | 5,634 | 1,473 |
| Terminal-Bench | — | 47.6% |
| APEX-Agents | 40.9% | — |
| DeepResearch Bench | — | 42.8% |
| ExploitBench | 18.1% | — |
| GBAEval | 0% | — |
| LMArena Search | — | 1199 |
Reasoning Too close to call
GLM-5.1: 39.1 (#60), GPT-5.1: 39.8 (#58)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| SimpleBench | 55.1% | 53.2% |
| CritPt | 4.6% | 4.9% |
| Chess Puzzles | 19% | 32% |
| LMArena Hard Prompts | 1472 | 1457 |
| Epoch Capabilities Index | 149.84 | 149.64 |
| ARC-AGI-2 | — | 17.6% |
| NYT Connections (extended) | 77.7% | — |
| ARC-AGI-1 | — | 72.8% |
| EnigmaEval | — | 11.2% |
| Thematic Generalization | 69.8% | — |
| LiveBench Reasoning | — | 95.8% |
| Mystery Game Puzzles | — | 19% |
| DTBench | — | 90.1% |
| LiveBench Data Analysis | — | 72.1% |
| LMCA | — | 43.9% |
| ForecastBench | — | 58.1 |
| LiveBench | — | 78.8% |
Math GPT-5.1 leads
GLM-5.1: 49.7 (#60), GPT-5.1: 52.2 (#51)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 93.3% | 88.6% |
| LMArena Math | 1473 | 1447 |
| FrontierMath (Feb 2025 set) | 33.4% | 31% |
| FrontierMath Tier 4 (v1) | 12.5% | 12.5% |
| FrontierMath (Tiers 1-3) | 36.8% | — |
| MathArena Final-Answer Competitions | 67.1% | — |
| ProofBench | 22.2% | — |
| Omni-MATH | — | 46.4% |
| LiveBench Math | — | 94.5% |
Knowledge GLM-5.1 leads
GLM-5.1: 54.9 (#50), GPT-5.1: 50.6 (#71)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| GPQA Diamond | 89.9% | 87.6% |
| SimpleQA Verified | 34% | 48% |
| LMArena Expert | 1476 | 1470 |
| Humanity's Last Exam | — | 23.7% |
| MMLU-Pro | — | 57.9% |
| Vectara Hallucination Rate | — | 10.9% |
| GPQA (HELM) | — | 44.2% |
Multimodal Not comparable
GLM-5.1: —, GPT-5.1: 44.8 (#19)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| LMArena Vision | — | 1250 |
| VPCT | — | 58.7% |
| LMArena Document | — | 1403 |
Multilingual GLM-5.1 leads
GLM-5.1: 55.0 (#36), GPT-5.1: 53.8 (#56)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| LMArena Non-English | 1447 | 1431 |
| LMArena Chinese | 1515 | 1495 |
| LMArena French | 1474 | 1450 |
| LMArena German | 1465 | 1438 |
| LMArena Japanese | 1434 | 1453 |
| LMArena Korean | 1418 | 1401 |
| LMArena Russian | 1454 | 1435 |
| LMArena Spanish | 1469 | 1433 |
Instruction Following GPT-5.1 leads
GLM-5.1: 76.3 (#42), GPT-5.1: 83.9 (#1)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| LMArena Instruction Following | 1451 | 1443 |
| LiveBench Instruction Following | — | 93.3% |
| IFEval | — | 93.5% |
Long Context GPT-5.1 leads
GLM-5.1: 44.9 (#53), GPT-5.1: 47.6 (#14)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| LMArena Longer Query | 1466 | 1447 |
| CL-bench | — | 23.7% |
| CL-bench Life | — | 17.3% |
Writing & Preference GLM-5.1 leads
GLM-5.1: 66.9 (#31), GPT-5.1: 64.5 (#55)
| Benchmark | GLM-5.1 | GPT-5.1 |
|---|---|---|
| LMArena Text | 1461 | 1443 |
| LMArena Creative Writing | 1453 | 1427 |
| LMArena Multi-Turn | 1472 | 1450 |
| EQ-Bench Creative Writing | 1592 | — |
| WildBench | — | 86.3% |
| LiveBench Language | — | 80.2% |
Frequently asked questions
Is GLM-5.1 better than GPT-5.1?
GPT-5.1 is the stronger model overall, scoring 49.0 to 47.8 on the Noometry Index. GLM-5.1 costs 1.6× less per token, which makes it the better buy when GPT-5.1's lead doesn't matter for your workload.
Which is cheaper, GLM-5.1 or GPT-5.1?
GLM-5.1 is cheaper. It lists at $1.40 per million input tokens and $4.40 per million output tokens; GPT-5.1 lists at $1.25 and $10.
Is GLM-5.1 or GPT-5.1 better for coding?
GLM-5.1 scores higher on coding benchmarks: 48.7 versus 46.4 in the Noometry coding category.
Which has the bigger context window?
GPT-5.1 does, with 400K tokens against 200K.
How many benchmarks do GLM-5.1 and GPT-5.1 share?
32 benchmarks have published results for both models. GLM-5.1 has 41 scored results on Noometry and GPT-5.1 has 63.