Model comparison
GLM-5 vs GLM-5.3-Flash
GLM-5.3-Flash is the stronger model overall, scoring 49.6 to 45.9 on the Noometry Index.
Last verified . 23 shared benchmarks.
Summary
- They share 23 benchmarks with published results for both. GLM-5 scores higher in 1 category and GLM-5.3-Flash in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-5.3-Flash leads 53.1 to 44.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 80% for GLM-5 and 93.9% for GLM-5.3-Flash.
- GLM-5.3-Flash is cheaper at $0.15 / $0.50 per million input/output tokens, against $1 / $3.20 for GLM-5.
- GLM-5.3-Flash accepts more context: 1M tokens versus 205K.
Side by side
| GLM-5 | GLM-5.3-Flash | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Z.ai (Zhipu) |
| Noometry Index | 45.9 | 49.6 |
| Released | 2026-02-11 | 2026-08-20 |
| Weights | Open | Open |
| Context window | 205K | 1M |
| Max output | 131K | 131K |
| Input $ / M tokens | $1 | $0.15 |
| Output $ / M tokens | $3.20 | $0.50 |
| Results tracked | 45 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3-Flash leads
GLM-5: 49.0 (#47), GLM-5.3-Flash: 53.7 (#26)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| LMArena WebDev | 1434 | 1614 |
| LMArena Coding | 1461 | 1506 |
| ALE-Bench | 765.62 | 303.55 |
| SWE-bench Verified | 72.1% | — |
| DeepSWE | — | 63.4% |
| FrontierCode | — | 31.8% |
| SWE-bench Verified (bash only) | 72.8% | — |
| SWE-bench Multilingual | 69.7% | — |
| SciCode | — | 51.6% |
| WeirdML | 48.2% | — |
Agentic & Tool Use GLM-5.3-Flash leads
GLM-5: 30.9 (#65), GLM-5.3-Flash: 34.3 (#41)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| Terminal-Bench | 52.4% | — |
| APEX-Agents | — | 52.8% |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Banking | 9.8% | — |
| τ²-bench Retail | 73.7% | — |
| τ²-bench Telecom | 86.8% | — |
| GDP.pdf | — | 14% |
| Vending-Bench 2 | 4,432 | — |
Reasoning GLM-5.3-Flash leads
GLM-5: 27.6 (#100), GLM-5.3-Flash: 31.7 (#69)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| Chess Puzzles | 10% | 14% |
| LMArena Hard Prompts | 1451 | 1490 |
| Epoch Capabilities Index | 145.84 | 151.9 |
| ARC-AGI-2 | 4.9% | — |
| SimpleBench | 53.2% | — |
| Kagi LLM Benchmark | 75% | — |
| NYT Connections (extended) | 74.8% | — |
| ARC-AGI-1 | 44.7% | — |
| CritPt | — | 15.4% |
| Mystery Game Puzzles | — | 8% |
| Surface Evolver Bench | — | 52.5% |
| Bench to the Future 3 | — | 0.15 |
| ForecastBench | 61 | — |
Math GLM-5.3-Flash leads
GLM-5: 44.8 (#66), GLM-5.3-Flash: 53.1 (#38)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 80% | 93.9% |
| LMArena Math | 1440 | 1499 |
| FrontierMath (Tiers 1-3) | — | 55.8% |
| FrontierMath Tier 4 | — | 17.1% |
| MathArena Final-Answer Competitions | 65.7% | — |
| ProofBench | — | 21% |
| FrontierMath (Feb 2025 set) | 16.4% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GLM-5.3-Flash leads
GLM-5: 52.2 (#59), GLM-5.3-Flash: 58.4 (#33)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| GPQA Diamond | 87.8% | 90.2% |
| LMArena Expert | 1454 | 1512 |
| Vectara Hallucination Rate | 10.1% | — |
Multimodal Not comparable
GLM-5: —, GLM-5.3-Flash: 42.6 (#25)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| LMArena Vision | — | 1304 |
Multilingual GLM-5.3-Flash leads
GLM-5: 53.9 (#57), GLM-5.3-Flash: 56.2 (#24)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| LMArena Non-English | 1430 | 1462 |
| LMArena Chinese | 1511 | 1529 |
| LMArena French | 1455 | 1495 |
| LMArena German | 1445 | 1457 |
| LMArena Japanese | 1416 | 1420 |
| LMArena Korean | 1421 | 1443 |
| LMArena Russian | 1436 | 1472 |
| LMArena Spanish | 1452 | 1471 |
Instruction Following GLM-5.3-Flash leads
GLM-5: 75.4 (#64), GLM-5.3-Flash: 77.8 (#17)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| LMArena Instruction Following | 1427 | 1480 |
Long Context Too close to call
GLM-5: 44.6 (#58), GLM-5.3-Flash: 45.3 (#41)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| LMArena Longer Query | 1446 | 1479 |
| CL-bench | 18.7% | — |
Writing & Preference Too close to call
GLM-5: 66.2 (#38), GLM-5.3-Flash: 65.4 (#49)
| Benchmark | GLM-5 | GLM-5.3-Flash |
|---|---|---|
| LMArena Text | 1446 | 1471 |
| LMArena Creative Writing | 1438 | 1440 |
| LMArena Multi-Turn | 1456 | 1471 |
| EQ-Bench Creative Writing | 1601 | — |
Frequently asked questions
Is GLM-5 better than GLM-5.3-Flash?
GLM-5.3-Flash is the stronger model overall, scoring 49.6 to 45.9 on the Noometry Index.
Which is cheaper, GLM-5 or GLM-5.3-Flash?
GLM-5.3-Flash is cheaper. It lists at $0.15 per million input tokens and $0.50 per million output tokens; GLM-5 lists at $1 and $3.20.
Is GLM-5 or GLM-5.3-Flash better for coding?
GLM-5.3-Flash scores higher on coding benchmarks: 53.7 versus 49.0 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3-Flash does, with 1M tokens against 205K.
How many benchmarks do GLM-5 and GLM-5.3-Flash share?
23 benchmarks have published results for both models. GLM-5 has 45 scored results on Noometry and GLM-5.3-Flash has 36.