Model comparison
GLM-5.2 vs Llama 3.1-8B
GLM-5.2 is the stronger model overall, scoring 50.9 to 22.8 on the Noometry Index. Llama 3.1-8B costs 37× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
MetaSummary
- They share 27 benchmarks with published results for both. GLM-5.2 scores higher in 9 categories and Llama 3.1-8B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5.2 leads 57.1 to 8.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 86.4% for GLM-5.2 and 1.7% for Llama 3.1-8B.
- Llama 3.1-8B is cheaper at $0.05 / $0.08 per million input/output tokens, against $1.40 / $4.40 for GLM-5.2.
- GLM-5.2 accepts more context: 1M tokens versus 128K.
Side by side
| GLM-5.2 | Llama 3.1-8B | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Meta |
| Noometry Index | 50.9 | 22.8 |
| Released | 2026-06-13 | 2024-07-23 |
| Weights | Open | Open |
| Context window | 1M | 128K |
| Max output | 131K | 4K |
| Input $ / M tokens | $1.40 | $0.05 |
| Output $ / M tokens | $4.40 | $0.08 |
| Results tracked | 50 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.2 leads
GLM-5.2: 50.9 (#37), Llama 3.1-8B: 19.9 (#335)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| SciCode | 50.5% | 13.2% |
| WeirdML | 70.1% | 1.7% |
| LMArena Coding | 1484 | 1195 |
| SWE-bench Verified | 78.7% | — |
| DeepSWE | 43.8% | — |
| FrontierCode | 24.5% | — |
| LMArena WebDev | 1604 | — |
| BigCodeBench Instruct | — | 32.8% |
| BigCodeBench Complete | — | 40.5% |
| ALE-Bench | 1,047 | — |
| HumanEval+ | — | 62.8% |
| MBPP+ | — | 55.6% |
Agentic & Tool Use GLM-5.2 leads
GLM-5.2: 31.6 (#63), Llama 3.1-8B: 22.3 (#123)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 25.8% |
| τ²-bench Banking | 37.1% | — |
| PostTrainBench | 31.7% | — |
| BALROG | — | 15.1% |
| GBAEval | 0% | — |
| Vending-Bench 2 | 8,314 | — |
Reasoning GLM-5.2 leads
GLM-5.2: 42.1 (#46), Llama 3.1-8B: 14.5 (#316)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| CritPt | 20.9% | 0% |
| Chess Puzzles | 21% | 0% |
| LMArena Hard Prompts | 1480 | 1175 |
| DTBench | 93.6% | 50.9% |
| LMCA | 45.8% | 5.4% |
| Epoch Capabilities Index | 151.76 | 116.48 |
| ARC-AGI-2 | 22.8% | — |
| SimpleBench | 58.8% | — |
| Kagi LLM Benchmark | 62.6% | — |
| NYT Connections (extended) | 74.3% | — |
| ARC-AGI-1 | 77% | — |
| EBR-Bench | 9.5% | — |
| Mystery Game Puzzles | 19% | — |
| Surface Evolver Bench | 55.6% | — |
| PIQA | — | 81.2% |
Math GLM-5.2 leads
GLM-5.2: 55.4 (#35), Llama 3.1-8B: 9.5 (#311)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.4% | 1.7% |
| LMArena Math | 1482 | 1179 |
| FrontierMath (Tiers 1-3) | 59.2% | — |
| FrontierMath Tier 4 | 29.3% | — |
| MathArena Final-Answer Competitions | 67.6% | — |
| ProofBench | 35% | — |
| Omni-MATH | — | 13.7% |
| MATH Level 5 | — | 22.9% |
| GSM8K | — | 82.4% |
Knowledge GLM-5.2 leads
GLM-5.2: 57.1 (#36), Llama 3.1-8B: 8.0 (#302)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| GPQA Diamond | 91.9% | 27% |
| LMArena Expert | 1484 | 1144 |
| SimpleQA Verified | 34.2% | — |
| MMLU-Pro | — | 40.6% |
| GPQA (HELM) | — | 24.7% |
| BoolQ | — | 82.8% |
| MMLU | — | 56.1% |
Multilingual GLM-5.2 leads
GLM-5.2: 56.0 (#25), Llama 3.1-8B: 34.2 (#244)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| LMArena Non-English | 1459 | 1148 |
| LMArena Chinese | 1518 | 1151 |
| LMArena French | 1476 | 1176 |
| LMArena German | 1470 | 1144 |
| LMArena Japanese | 1451 | 1061 |
| LMArena Korean | 1445 | 1053 |
| LMArena Russian | 1467 | 1158 |
| LMArena Spanish | 1477 | 1169 |
Instruction Following GLM-5.2 leads
GLM-5.2: 77.1 (#32), Llama 3.1-8B: 59.1 (#253)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| LMArena Instruction Following | 1465 | 1159 |
| IFEval | — | 74.3% |
Long Context GLM-5.2 leads
GLM-5.2: 45.3 (#43), Llama 3.1-8B: 35.8 (#233)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| LMArena Longer Query | 1479 | 1182 |
Writing & Preference GLM-5.2 leads
GLM-5.2: 70.7 (#20), Llama 3.1-8B: 29.9 (#285)
| Benchmark | GLM-5.2 | Llama 3.1-8B |
|---|---|---|
| LMArena Text | 1470 | 1187 |
| LMArena Creative Writing | 1465 | 1154 |
| EQ-Bench Creative Writing | 1757 | 713 |
| LMArena Multi-Turn | 1468 | 1172 |
| WildBench | — | 68.7% |
| EQ-Bench 4 | 1222 | — |
Frequently asked questions
Is GLM-5.2 better than Llama 3.1-8B?
GLM-5.2 is the stronger model overall, scoring 50.9 to 22.8 on the Noometry Index. Llama 3.1-8B costs 37× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.
Which is cheaper, GLM-5.2 or Llama 3.1-8B?
Llama 3.1-8B is cheaper. It lists at $0.05 per million input tokens and $0.08 per million output tokens; GLM-5.2 lists at $1.40 and $4.40.
Is GLM-5.2 or Llama 3.1-8B better for coding?
GLM-5.2 scores higher on coding benchmarks: 50.9 versus 19.9 in the Noometry coding category.
Which has the bigger context window?
GLM-5.2 does, with 1M tokens against 128K.
How many benchmarks do GLM-5.2 and Llama 3.1-8B share?
27 benchmarks have published results for both models. GLM-5.2 has 50 scored results on Noometry and Llama 3.1-8B has 43.