Model comparison
GLM-4.5-Air vs o3
o3 is the stronger model overall, scoring 47.5 to 38.9 on the Noometry Index. GLM-4.5-Air costs 8.2× less per token, which makes it the better buy when o3's lead doesn't matter for your workload.
Last verified . 26 shared benchmarks.
Summary
- They share 26 benchmarks with published results for both. GLM-4.5-Air scores higher in 0 categories and o3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where o3 leads 54.6 to 35.0.
- The biggest single-benchmark swing is Omni-MATH: 39.1% for GLM-4.5-Air and 71.4% for o3.
- GLM-4.5-Air is cheaper at $0.20 / $1.10 per million input/output tokens, against $2 / $8 for o3.
- o3 accepts more context: 200K tokens versus 131K.
- GLM-4.5-Air has downloadable open weights; the other is API-only.
Side by side
| GLM-4.5-Air | o3 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 38.9 | 47.5 |
| Released | 2025-07-20 | 2025-04-16 |
| Weights | Open | Proprietary |
| Context window | 131K | 200K |
| Max output | 98K | 100K |
| Input $ / M tokens | $0.20 | $2 |
| Output $ / M tokens | $1.10 | $8 |
| Results tracked | 27 | 63 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o3 leads
GLM-4.5-Air: 33.3 (#259), o3: 46.8 (#64)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| GSO | 2.9% | 8.8% |
| LMArena Coding | 1397 | 1408 |
| SWE-bench Verified | — | 62.3% |
| SWE-bench Verified (bash only) | — | 58.4% |
| Aider Polyglot | — | 81.3% |
| WeirdML | — | 52.4% |
| CadEval | — | 74% |
| ALE-Bench | — | 933.55 |
Agentic & Tool Use Not comparable
GLM-4.5-Air: —, o3: 34.5 (#44)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 30.8% |
| DeepResearch Bench | — | 45.2% |
| OSWorld | — | 23% |
| LMArena Search | — | 1144 |
| METR Time Horizons | — | 65.4% |
Reasoning o3 leads
GLM-4.5-Air: 24.1 (#166), o3: 32.0 (#78)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| Kagi LLM Benchmark | 43% | 67.6% |
| LMArena Hard Prompts | 1379 | 1402 |
| ForecastBench | 59.2 | 62.5 |
| ARC-AGI-2 | — | 6.5% |
| SimpleBench | — | 53.1% |
| ARC-AGI-1 | — | 60.8% |
| CritPt | — | 1.4% |
| Chess Puzzles | — | 38% |
| EnigmaEval | — | 13.1% |
| Mystery Game Puzzles | — | 29% |
| DTBench | — | 84.8% |
| LMCA | — | 39.7% |
| Epoch Capabilities Index | — | 146.86 |
Math o3 leads
GLM-4.5-Air: 36.2 (#170), o3: 50.2 (#58)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| Omni-MATH | 39.1% | 71.4% |
| LMArena Math | 1396 | 1426 |
| FrontierMath (Tiers 1-3) | — | 33.3% |
| OTIS Mock AIME 2024-2025 | — | 84.4% |
| MATH Level 5 | — | 97.8% |
| FrontierMath (Feb 2025 set) | — | 18.7% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge o3 leads
GLM-4.5-Air: 35.0 (#191), o3: 54.6 (#52)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| Humanity's Last Exam | 8.1% | 20.3% |
| MMLU-Pro | 76.2% | 85.9% |
| GPQA (HELM) | 59.4% | 75.3% |
| LMArena Expert | 1370 | 1402 |
| GPQA Diamond | — | 81.8% |
| SimpleQA Verified | — | 49.4% |
| Confabulations | — | 14.4% |
| Vectara Hallucination Rate | 9.3% | — |
Multimodal Not comparable
GLM-4.5-Air: —, o3: 41.4 (#36)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| LMArena Vision | — | 1214 |
| GeoBench | — | 74% |
| VPCT | — | 52% |
Multilingual o3 leads
GLM-4.5-Air: 49.1 (#135), o3: 51.7 (#105)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| LMArena Non-English | 1366 | 1401 |
| LMArena Chinese | 1426 | 1437 |
| LMArena French | 1399 | 1430 |
| LMArena German | 1377 | 1420 |
| LMArena Japanese | 1348 | 1403 |
| LMArena Korean | 1308 | 1370 |
| LMArena Russian | 1373 | 1406 |
| LMArena Spanish | 1386 | 1395 |
Instruction Following o3 leads
GLM-4.5-Air: 69.6 (#171), o3: 72.8 (#127)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| IFEval | 81.2% | 86.9% |
| LMArena Instruction Following | 1354 | 1368 |
Long Context o3 leads
GLM-4.5-Air: 41.6 (#135), o3: 53.3 (#6)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| LMArena Longer Query | 1366 | 1372 |
| Fiction.LiveBench | — | 88.9% |
| CL-bench | — | 17.8% |
Writing & Preference o3 leads
GLM-4.5-Air: 55.9 (#139), o3: 63.5 (#64)
| Benchmark | GLM-4.5-Air | o3 |
|---|---|---|
| LMArena Text | 1384 | 1410 |
| LMArena Creative Writing | 1343 | 1359 |
| WildBench | 78.9% | 86.1% |
| LMArena Multi-Turn | 1371 | 1405 |
| Short-Story Creative Writing | — | 83.9% |
| EQ-Bench Creative Writing | — | 1676 |
Frequently asked questions
Is GLM-4.5-Air better than o3?
o3 is the stronger model overall, scoring 47.5 to 38.9 on the Noometry Index. GLM-4.5-Air costs 8.2× less per token, which makes it the better buy when o3's lead doesn't matter for your workload.
Which is cheaper, GLM-4.5-Air or o3?
GLM-4.5-Air is cheaper. It lists at $0.20 per million input tokens and $1.10 per million output tokens; o3 lists at $2 and $8.
Is GLM-4.5-Air or o3 better for coding?
o3 scores higher on coding benchmarks: 46.8 versus 33.3 in the Noometry coding category.
Which has the bigger context window?
o3 does, with 200K tokens against 131K.
How many benchmarks do GLM-4.5-Air and o3 share?
26 benchmarks have published results for both models. GLM-4.5-Air has 27 scored results on Noometry and o3 has 63.