Model comparison
GLM-4.7 vs Llama-3.3-70B-Instruct
GLM-4.7 is the stronger model overall, scoring 42.0 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 6.5× less per token, which makes it the better buy when GLM-4.7's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
MetaSummary
- They share 24 benchmarks with published results for both. GLM-4.7 scores higher in 9 categories and Llama-3.3-70B-Instruct in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-4.7 leads 38.6 to 15.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 83.3% for GLM-4.7 and 5.1% for Llama-3.3-70B-Instruct.
- Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $0.60 / $2.20 for GLM-4.7.
- GLM-4.7 accepts more context: 205K tokens versus 128K.
Side by side
| GLM-4.7 | Llama-3.3-70B-Instruct | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Meta |
| Noometry Index | 42.0 | 30.6 |
| Released | 2025-12-22 | 2024-12-06 |
| Weights | Open | Open |
| Context window | 205K | 128K |
| Max output | 131K | 4K |
| Input $ / M tokens | $0.60 | $0.10 |
| Output $ / M tokens | $2.20 | $0.32 |
| Results tracked | 36 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-4.7 leads
GLM-4.7: 44.0 (#79), Llama-3.3-70B-Instruct: 31.0 (#290)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| SciCode | 45.1% | 26% |
| LMArena Coding | 1454 | 1268 |
| LMArena WebDev | 1435 | — |
| WeirdML | — | 14.4% |
| BigCodeBench Instruct | — | 46.9% |
| LiveBench Coding | — | 36.6% |
| BigCodeBench Complete | — | 57.5% |
| ALE-Bench | 399.48 | — |
Agentic & Tool Use Too close to call
GLM-4.7: 26.5 (#103), Llama-3.3-70B-Instruct: 25.8 (#105)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| Terminal-Bench | 33.4% | — |
| Berkeley Function Calling Leaderboard | — | 31.9% |
| BALROG | — | 23% |
| Vending-Bench 2 | 2,377 | — |
Reasoning GLM-4.7 leads
GLM-4.7: 24.3 (#164), Llama-3.3-70B-Instruct: 14.1 (#327)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| SimpleBench | 47.7% | 19.9% |
| CritPt | 1.7% | 0% |
| LMArena Hard Prompts | 1443 | 1257 |
| Epoch Capabilities Index | 143.51 | 127.33 |
| Chess Puzzles | 6% | — |
| LiveBench Reasoning | — | 50.8% |
| DTBench | — | 59.5% |
| LiveBench Data Analysis | — | 49.5% |
| LMCA | — | 17.5% |
| ForecastBench | — | 58.6 |
| LiveBench | — | 50.2% |
Math GLM-4.7 leads
GLM-4.7: 38.6 (#135), Llama-3.3-70B-Instruct: 15.3 (#298)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 83.3% | 5.1% |
| LMArena Math | 1423 | 1267 |
| ProofBench | 6% | — |
| LiveBench Math | — | 42.2% |
| MATH Level 5 | — | 41.6% |
| FrontierMath (Feb 2025 set) | 2.4% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge GLM-4.7 leads
GLM-4.7: 47.0 (#80), Llama-3.3-70B-Instruct: 30.6 (#226)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| GPQA Diamond | 83.3% | 47.4% |
| Vectara Hallucination Rate | 11.7% | 4.1% |
| LMArena Expert | 1424 | 1225 |
| SimpleQA Verified | 32.2% | — |
| Confabulations | — | 22.8% |
| MMLU | — | 86.3% |
Multilingual GLM-4.7 leads
GLM-4.7: 52.8 (#79), Llama-3.3-70B-Instruct: 39.9 (#220)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1417 | 1236 |
| LMArena Chinese | 1495 | 1217 |
| LMArena French | 1432 | 1281 |
| LMArena German | 1424 | 1251 |
| LMArena Japanese | 1439 | 1150 |
| LMArena Korean | 1399 | 1143 |
| LMArena Russian | 1423 | 1252 |
| LMArena Spanish | 1434 | 1270 |
Instruction Following GLM-4.7 leads
GLM-4.7: 74.4 (#95), Llama-3.3-70B-Instruct: 71.1 (#157)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Instruction Following | 1411 | 1242 |
| LiveBench Instruction Following | — | 82.7% |
Long Context GLM-4.7 leads
GLM-4.7: 42.8 (#116), Llama-3.3-70B-Instruct: 26.4 (#295)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Longer Query | 1432 | 1256 |
| Fiction.LiveBench | — | 33.3% |
| CL-bench | 15.9% | — |
| CL-bench Life | 10.9% | — |
Writing & Preference GLM-4.7 leads
GLM-4.7: 60.9 (#93), Llama-3.3-70B-Instruct: 47.6 (#207)
| Benchmark | GLM-4.7 | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1435 | 1274 |
| LMArena Creative Writing | 1401 | 1250 |
| LMArena Multi-Turn | 1446 | 1280 |
| EQ-Bench Creative Writing | 1413 | — |
| LiveBench Language | — | 39.2% |
Frequently asked questions
Is GLM-4.7 better than Llama-3.3-70B-Instruct?
GLM-4.7 is the stronger model overall, scoring 42.0 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 6.5× less per token, which makes it the better buy when GLM-4.7's lead doesn't matter for your workload.
Which is cheaper, GLM-4.7 or Llama-3.3-70B-Instruct?
Llama-3.3-70B-Instruct is cheaper. It lists at $0.10 per million input tokens and $0.32 per million output tokens; GLM-4.7 lists at $0.60 and $2.20.
Is GLM-4.7 or Llama-3.3-70B-Instruct better for coding?
GLM-4.7 scores higher on coding benchmarks: 44.0 versus 31.0 in the Noometry coding category.
Which has the bigger context window?
GLM-4.7 does, with 205K tokens against 128K.
How many benchmarks do GLM-4.7 and Llama-3.3-70B-Instruct share?
24 benchmarks have published results for both models. GLM-4.7 has 36 scored results on Noometry and Llama-3.3-70B-Instruct has 43.