Model comparison
GLM-5 vs Llama 3.1-8B
GLM-5 is the stronger model overall, scoring 46.1 to 23.0 on the Noometry Index. Llama 3.1-8B costs 27× less per token, which makes it the better buy when GLM-5's lead doesn't matter for your workload.
Last verified . 23 shared benchmarks.
MetaSummary
- They share 23 benchmarks with published results for both. GLM-5 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 leads 52.3 to 8.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 80% for GLM-5 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 / $3.20 for GLM-5.
- GLM-5 accepts more context: 205K tokens versus 128K.
Side by side
| GLM-5 | Llama 3.1-8B | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Meta |
| Noometry Index | 46.1 | 23.0 |
| Released | 2026-02-11 | 2024-07-23 |
| Weights | Open | Open |
| Context window | 205K | 128K |
| Max output | 131K | 4K |
| Input $ / M tokens | $1 | $0.05 |
| Output $ / M tokens | $3.20 | $0.08 |
| Results tracked | 45 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5 leads
GLM-5: 49.0 (#52), Llama 3.1-8B: 20.2 (#340)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| WeirdML | 48.2% | 1.7% |
| LMArena Coding | 1461 | 1195 |
| SWE-bench Verified | 72.1% | — |
| SWE-bench Verified (bash only) | 72.8% | — |
| LMArena WebDev | 1434 | — |
| SWE-bench Multilingual | 69.7% | — |
| SciCode | — | 13.2% |
| BigCodeBench Instruct | — | 32.8% |
| BigCodeBench Complete | — | 40.5% |
| ALE-Bench | 765.62 | — |
| HumanEval+ | — | 62.8% |
| MBPP+ | — | 55.6% |
Agentic & Tool Use GLM-5 leads
GLM-5: 31.1 (#71), Llama 3.1-8B: 22.5 (#131)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| Terminal-Bench | 52.4% | — |
| Berkeley Function Calling Leaderboard | — | 25.8% |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Banking | 9.8% | — |
| τ²-bench Retail | 73.7% | — |
| τ²-bench Telecom | 86.8% | — |
| BALROG | — | 15.1% |
| Vending-Bench 2 | 4,432 | — |
Reasoning GLM-5 leads
GLM-5: 27.6 (#116), Llama 3.1-8B: 14.9 (#321)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| Chess Puzzles | 10% | 0% |
| LMArena Hard Prompts | 1452 | 1175 |
| Epoch Capabilities Index | 145.83 | 116.57 |
| ARC-AGI-2 | 4.9% | — |
| SimpleBench | 53.2% | — |
| Kagi LLM Benchmark | 75% | — |
| NYT Connections (extended) | 74.8% | — |
| ARC-AGI-1 | 44.7% | — |
| CritPt | — | 0% |
| DTBench | — | 50.9% |
| LMCA | — | 5.4% |
| ForecastBench | 61 | — |
| PIQA | — | 81.2% |
Math GLM-5 leads
GLM-5: 46.4 (#71), Llama 3.1-8B: 10.2 (#317)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 80% | 1.7% |
| LMArena Math | 1440 | 1179 |
| MathArena Final-Answer Competitions | 65.7% | — |
| Omni-MATH | — | 13.7% |
| MATH Level 5 | — | 22.9% |
| FrontierMath (Feb 2025 set) | 16.4% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
| GSM8K | — | 82.4% |
Knowledge GLM-5 leads
GLM-5: 52.3 (#64), Llama 3.1-8B: 8.0 (#307)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| GPQA Diamond | 87.8% | 27% |
| LMArena Expert | 1454 | 1144 |
| MMLU-Pro | — | 40.6% |
| Vectara Hallucination Rate | 10.1% | — |
| GPQA (HELM) | — | 24.7% |
| BoolQ | — | 82.8% |
| MMLU | — | 56.1% |
Multilingual GLM-5 leads
GLM-5: 53.7 (#58), Llama 3.1-8B: 34.0 (#249)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| LMArena Non-English | 1430 | 1148 |
| LMArena Chinese | 1511 | 1151 |
| LMArena French | 1455 | 1177 |
| LMArena German | 1445 | 1144 |
| LMArena Japanese | 1416 | 1061 |
| LMArena Korean | 1423 | 1053 |
| LMArena Russian | 1436 | 1158 |
| LMArena Spanish | 1454 | 1169 |
Instruction Following GLM-5 leads
GLM-5: 75.2 (#67), Llama 3.1-8B: 58.9 (#258)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| LMArena Instruction Following | 1428 | 1159 |
| IFEval | — | 74.3% |
Long Context GLM-5 leads
GLM-5: 44.7 (#60), Llama 3.1-8B: 35.8 (#238)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| LMArena Longer Query | 1446 | 1182 |
| CL-bench | 18.7% | — |
Writing & Preference GLM-5 leads
GLM-5: 66.0 (#38), Llama 3.1-8B: 29.7 (#290)
| Benchmark | GLM-5 | Llama 3.1-8B |
|---|---|---|
| LMArena Text | 1446 | 1187 |
| LMArena Creative Writing | 1439 | 1154 |
| EQ-Bench Creative Writing | 1601 | 713 |
| LMArena Multi-Turn | 1456 | 1172 |
| WildBench | — | 68.7% |
Frequently asked questions
Is GLM-5 better than Llama 3.1-8B?
GLM-5 is the stronger model overall, scoring 46.1 to 23.0 on the Noometry Index. Llama 3.1-8B costs 27× less per token, which makes it the better buy when GLM-5's lead doesn't matter for your workload.
Which is cheaper, GLM-5 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 lists at $1 and $3.20.
Is GLM-5 or Llama 3.1-8B better for coding?
GLM-5 scores higher on coding benchmarks: 49.0 versus 20.2 in the Noometry coding category.
Which has the bigger context window?
GLM-5 does, with 205K tokens against 128K.
How many benchmarks do GLM-5 and Llama 3.1-8B share?
23 benchmarks have published results for both models. GLM-5 has 45 scored results on Noometry and Llama 3.1-8B has 43.