Model comparison
GLM-5.3 vs Llama 4 Scout
GLM-5.3 is the stronger model overall, scoring 54.8 to 27.7 on the Noometry Index. Llama 4 Scout costs 14× less per token, which makes it the better buy when GLM-5.3's lead doesn't matter for your workload.
Last verified . 25 shared benchmarks.
MetaSummary
- They share 25 benchmarks with published results for both. GLM-5.3 scores higher in 9 categories and Llama 4 Scout in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-5.3 leads 62.3 to 19.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 91.1% for GLM-5.3 and 7.8% for Llama 4 Scout.
- Llama 4 Scout is cheaper at $0.10 / $0.30 per million input/output tokens, against $1.40 / $4.40 for GLM-5.3.
- GLM-5.3 accepts more context: 1M tokens versus 128K.
Side by side
| GLM-5.3 | Llama 4 Scout | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Meta |
| Noometry Index | 54.8 | 27.7 |
| Released | 2026-08-14 | 2025-04-05 |
| Weights | Open | Open |
| Context window | 1M | 128K |
| Max output | 131K | 4K |
| Input $ / M tokens | $1.40 | $0.10 |
| Output $ / M tokens | $4.40 | $0.30 |
| Results tracked | 42 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3 leads
GLM-5.3: 59.5 (#14), Llama 4 Scout: 20.2 (#339)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| SciCode | 59% | 17% |
| LMArena Coding | 1496 | 1286 |
| DeepSWE | 69% | — |
| FrontierCode | 40.1% | — |
| SWE-bench Verified (bash only) | — | 9.1% |
| CursorBench | 42.6% | — |
| LMArena WebDev | 1622 | — |
| FrontierSWE | 30.2% | — |
| WeirdML | 75.4% | — |
| BigCodeBench Complete | — | 43.1% |
| ALE-Bench | 1,317 | — |
Agentic & Tool Use GLM-5.3 leads
GLM-5.3: 36.4 (#38), Llama 4 Scout: 24.6 (#119)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| APEX-Agents | 56.6% | — |
| Berkeley Function Calling Leaderboard | — | 28.1% |
| Vending-Bench 2 | 8,164 | — |
Reasoning GLM-5.3 leads
GLM-5.3: 46.1 (#46), Llama 4 Scout: 9.1 (#345)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| CritPt | 19.1% | 0% |
| LMArena Hard Prompts | 1489 | 1266 |
| DTBench | 87.7% | 57.9% |
| LMCA | 55.5% | 12% |
| Epoch Capabilities Index | 155.61 | 129.64 |
| ARC-AGI-2 | — | 0% |
| Kagi LLM Benchmark | — | 36.9% |
| NYT Connections (extended) | 74.2% | — |
| ARC-AGI-1 | — | 0.5% |
| Chess Puzzles | 21% | — |
| Mystery Game Puzzles | 33% | — |
| Bench to the Future 3 | 0.15 | — |
| ForecastBench | — | 57.5 |
Math GLM-5.3 leads
GLM-5.3: 62.3 (#33), Llama 4 Scout: 19.6 (#286)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.1% | 7.8% |
| LMArena Math | 1489 | 1287 |
| FrontierMath (Tiers 1-3) | 68.8% | — |
| FrontierMath Tier 4 | 29.3% | — |
| ProofBench | 49% | — |
| Omni-MATH | — | 37.3% |
| MATH Level 5 | — | 62.3% |
| FrontierMath (Feb 2025 set) | — | 0% |
Knowledge GLM-5.3 leads
GLM-5.3: 58.3 (#37), Llama 4 Scout: 31.9 (#217)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| GPQA Diamond | 90.9% | 51.8% |
| LMArena Expert | 1516 | 1235 |
| SimpleQA Verified | 41% | — |
| MMLU-Pro | — | 74.2% |
| Vectara Hallucination Rate | — | 7.7% |
| GPQA (HELM) | — | 50.7% |
Multimodal Not comparable
GLM-5.3: —, Llama 4 Scout: 32.2 (#102)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| LMArena Vision | — | 1118 |
| SpatialViz-Bench | — | 34.2% |
Multilingual GLM-5.3 leads
GLM-5.3: 55.7 (#28), Llama 4 Scout: 41.0 (#212)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| LMArena Non-English | 1457 | 1252 |
| LMArena Chinese | 1528 | 1255 |
| LMArena French | 1499 | 1282 |
| LMArena German | 1499 | 1272 |
| LMArena Japanese | 1453 | 1206 |
| LMArena Korean | 1472 | 1207 |
| LMArena Russian | 1463 | 1263 |
| LMArena Spanish | 1460 | 1278 |
Instruction Following GLM-5.3 leads
GLM-5.3: 77.5 (#23), Llama 4 Scout: 65.8 (#217)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| LMArena Instruction Following | 1477 | 1248 |
| IFEval | — | 81.8% |
Long Context GLM-5.3 leads
GLM-5.3: 45.4 (#41), Llama 4 Scout: 27.5 (#294)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| LMArena Longer Query | 1482 | 1265 |
| Fiction.LiveBench | — | 36% |
Writing & Preference GLM-5.3 leads
GLM-5.3: 75.7 (#6), Llama 4 Scout: 37.0 (#261)
| Benchmark | GLM-5.3 | Llama 4 Scout |
|---|---|---|
| LMArena Text | 1471 | 1279 |
| LMArena Creative Writing | 1457 | 1249 |
| EQ-Bench Creative Writing | 2075 | 783 |
| LMArena Multi-Turn | 1472 | 1280 |
| WildBench | — | 78% |
Frequently asked questions
Is GLM-5.3 better than Llama 4 Scout?
GLM-5.3 is the stronger model overall, scoring 54.8 to 27.7 on the Noometry Index. Llama 4 Scout costs 14× less per token, which makes it the better buy when GLM-5.3's lead doesn't matter for your workload.
Which is cheaper, GLM-5.3 or Llama 4 Scout?
Llama 4 Scout is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; GLM-5.3 lists at $1.40 and $4.40.
Is GLM-5.3 or Llama 4 Scout better for coding?
GLM-5.3 scores higher on coding benchmarks: 59.5 versus 20.2 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3 does, with 1M tokens against 128K.
How many benchmarks do GLM-5.3 and Llama 4 Scout share?
25 benchmarks have published results for both models. GLM-5.3 has 42 scored results on Noometry and Llama 4 Scout has 43.