Model comparison
GLM-4.7 vs o3
o3 is the stronger model overall, scoring 47.5 to 42.0 on the Noometry Index. GLM-4.7 costs 3.5× less per token, which makes it the better buy when o3's lead doesn't matter for your workload.
Last verified . 29 shared benchmarks.
Summary
- They share 29 benchmarks with published results for both. GLM-4.7 scores higher in 2 categories and o3 in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where o3 leads 50.2 to 38.6.
- The biggest single-benchmark swing is Chess Puzzles: 6% for GLM-4.7 and 38% for o3.
- GLM-4.7 is cheaper at $0.60 / $2.20 per million input/output tokens, against $2 / $8 for o3.
- GLM-4.7 accepts more context: 205K tokens versus 200K.
- GLM-4.7 has downloadable open weights; the other is API-only.
Side by side
| GLM-4.7 | o3 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 42.0 | 47.5 |
| Released | 2025-12-22 | 2025-04-16 |
| Weights | Open | Proprietary |
| Context window | 205K | 200K |
| Max output | 131K | 100K |
| Input $ / M tokens | $0.60 | $2 |
| Output $ / M tokens | $2.20 | $8 |
| Results tracked | 36 | 63 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o3 leads
GLM-4.7: 44.0 (#79), o3: 46.8 (#64)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| LMArena Coding | 1454 | 1408 |
| ALE-Bench | 399.48 | 933.55 |
| SWE-bench Verified | — | 62.3% |
| SWE-bench Verified (bash only) | — | 58.4% |
| Aider Polyglot | — | 81.3% |
| LMArena WebDev | 1435 | — |
| SciCode | 45.1% | — |
| GSO | — | 8.8% |
| WeirdML | — | 52.4% |
| CadEval | — | 74% |
Agentic & Tool Use o3 leads
GLM-4.7: 26.5 (#103), o3: 34.5 (#44)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| Terminal-Bench | 33.4% | — |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 30.8% |
| DeepResearch Bench | — | 45.2% |
| OSWorld | — | 23% |
| LMArena Search | — | 1144 |
| METR Time Horizons | — | 65.4% |
| Vending-Bench 2 | 2,377 | — |
Reasoning o3 leads
GLM-4.7: 24.3 (#164), o3: 32.0 (#78)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| SimpleBench | 47.7% | 53.1% |
| CritPt | 1.7% | 1.4% |
| Chess Puzzles | 6% | 38% |
| LMArena Hard Prompts | 1443 | 1402 |
| Epoch Capabilities Index | 143.51 | 146.86 |
| ARC-AGI-2 | — | 6.5% |
| Kagi LLM Benchmark | — | 67.6% |
| ARC-AGI-1 | — | 60.8% |
| EnigmaEval | — | 13.1% |
| Mystery Game Puzzles | — | 29% |
| DTBench | — | 84.8% |
| LMCA | — | 39.7% |
| ForecastBench | — | 62.5 |
Math o3 leads
GLM-4.7: 38.6 (#135), o3: 50.2 (#58)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 83.3% | 84.4% |
| LMArena Math | 1423 | 1426 |
| FrontierMath (Feb 2025 set) | 2.4% | 18.7% |
| FrontierMath Tier 4 (v1) | 0% | 2.1% |
| FrontierMath (Tiers 1-3) | — | 33.3% |
| ProofBench | 6% | — |
| Omni-MATH | — | 71.4% |
| MATH Level 5 | — | 97.8% |
Knowledge o3 leads
GLM-4.7: 47.0 (#80), o3: 54.6 (#52)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| GPQA Diamond | 83.3% | 81.8% |
| SimpleQA Verified | 32.2% | 49.4% |
| LMArena Expert | 1424 | 1402 |
| Humanity's Last Exam | — | 20.3% |
| MMLU-Pro | — | 85.9% |
| Confabulations | — | 14.4% |
| Vectara Hallucination Rate | 11.7% | — |
| GPQA (HELM) | — | 75.3% |
Multimodal Not comparable
GLM-4.7: —, o3: 41.4 (#36)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| LMArena Vision | — | 1214 |
| GeoBench | — | 74% |
| VPCT | — | 52% |
Multilingual GLM-4.7 leads
GLM-4.7: 52.8 (#79), o3: 51.7 (#105)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| LMArena Non-English | 1417 | 1401 |
| LMArena Chinese | 1495 | 1437 |
| LMArena French | 1432 | 1430 |
| LMArena German | 1424 | 1420 |
| LMArena Japanese | 1439 | 1403 |
| LMArena Korean | 1399 | 1370 |
| LMArena Russian | 1423 | 1406 |
| LMArena Spanish | 1434 | 1395 |
Instruction Following GLM-4.7 leads
GLM-4.7: 74.4 (#95), o3: 72.8 (#127)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| LMArena Instruction Following | 1411 | 1368 |
| IFEval | — | 86.9% |
Long Context o3 leads
GLM-4.7: 42.8 (#116), o3: 53.3 (#6)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| CL-bench | 15.9% | 17.8% |
| LMArena Longer Query | 1432 | 1372 |
| Fiction.LiveBench | — | 88.9% |
| CL-bench Life | 10.9% | — |
Writing & Preference o3 leads
GLM-4.7: 60.9 (#93), o3: 63.5 (#64)
| Benchmark | GLM-4.7 | o3 |
|---|---|---|
| LMArena Text | 1435 | 1410 |
| LMArena Creative Writing | 1401 | 1359 |
| EQ-Bench Creative Writing | 1413 | 1676 |
| LMArena Multi-Turn | 1446 | 1405 |
| Short-Story Creative Writing | — | 83.9% |
| WildBench | — | 86.1% |
Frequently asked questions
Is GLM-4.7 better than o3?
o3 is the stronger model overall, scoring 47.5 to 42.0 on the Noometry Index. GLM-4.7 costs 3.5× less per token, which makes it the better buy when o3's lead doesn't matter for your workload.
Which is cheaper, GLM-4.7 or o3?
GLM-4.7 is cheaper. It lists at $0.60 per million input tokens and $2.20 per million output tokens; o3 lists at $2 and $8.
Is GLM-4.7 or o3 better for coding?
o3 scores higher on coding benchmarks: 46.8 versus 44.0 in the Noometry coding category.
Which has the bigger context window?
GLM-4.7 does, with 205K tokens against 200K.
How many benchmarks do GLM-4.7 and o3 share?
29 benchmarks have published results for both models. GLM-4.7 has 36 scored results on Noometry and o3 has 63.