Model comparison
GLM-4.5-Air vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 38.9 on the Noometry Index. GLM-4.5-Air costs 14× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 18 shared benchmarks.
Summary
- They share 18 benchmarks with published results for both. GLM-4.5-Air scores higher in 0 categories and Kimi K3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K3 leads 63.0 to 24.1.
- GLM-4.5-Air is cheaper at $0.20 / $1.10 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 131K.
Side by side
| GLM-4.5-Air | Kimi K3 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Moonshot AI |
| Noometry Index | 38.9 | 59.5 |
| Released | 2025-07-20 | 2026-07-16 |
| Weights | Open | Open |
| Context window | 131K | 1.05M |
| Max output | 98K | 1.05M |
| Input $ / M tokens | $0.20 | $3 |
| Output $ / M tokens | $1.10 | $15 |
| Results tracked | 27 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
GLM-4.5-Air: 33.3 (#259), Kimi K3: 61.0 (#10)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| LMArena Coding | 1397 | 1508 |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| LMArena WebDev | — | 1654 |
| FrontierSWE | — | 25.9% |
| SciCode | — | 59.5% |
| GSO | 2.9% | — |
| WeirdML | — | 82.6% |
| ALE-Bench | — | 1,524 |
Agentic & Tool Use Not comparable
GLM-4.5-Air: —, Kimi K3: 41.8 (#20)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| APEX-Agents | — | 50.6% |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| Vending-Bench 2 | — | 5,165 |
Reasoning Kimi K3 leads
GLM-4.5-Air: 24.1 (#166), Kimi K3: 63.0 (#17)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| LMArena Hard Prompts | 1379 | 1496 |
| ForecastBench | 59.2 | 61.1 |
| ARC-AGI-2 | — | 60.4% |
| SimpleBench | — | 60.7% |
| Kagi LLM Benchmark | 43% | — |
| NYT Connections (extended) | — | 93.6% |
| ARC-AGI-1 | — | 94.5% |
| CritPt | — | 23.4% |
| Chess Puzzles | — | 39% |
| Mystery Game Puzzles | — | 26% |
| DTBench | — | 91.2% |
| LMCA | — | 52.7% |
| Surface Evolver Bench | — | 95% |
| Epoch Capabilities Index | — | 157.45 |
Math Kimi K3 leads
GLM-4.5-Air: 36.2 (#170), Kimi K3: 74.2 (#16)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| LMArena Math | 1396 | 1491 |
| FrontierMath (Tiers 1-3) | — | 72.2% |
| FrontierMath Tier 4 | — | 39% |
| MathArena Final-Answer Competitions | — | 87.8% |
| OTIS Mock AIME 2024-2025 | — | 97.2% |
| ProofBench | — | 87% |
| Omni-MATH | 39.1% | — |
Knowledge Kimi K3 leads
GLM-4.5-Air: 35.0 (#191), Kimi K3: 63.2 (#21)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| LMArena Expert | 1370 | 1521 |
| GPQA Diamond | — | 93.1% |
| Humanity's Last Exam | 8.1% | — |
| SimpleQA Verified | — | 50.6% |
| MMLU-Pro | 76.2% | — |
| Vectara Hallucination Rate | 9.3% | — |
| GPQA (HELM) | 59.4% | — |
Multimodal Not comparable
GLM-4.5-Air: —, Kimi K3: 37.8 (#70)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
Multilingual Kimi K3 leads
GLM-4.5-Air: 49.1 (#135), Kimi K3: 56.3 (#21)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1366 | 1466 |
| LMArena Chinese | 1426 | 1529 |
| LMArena French | 1399 | 1491 |
| LMArena German | 1377 | 1488 |
| LMArena Japanese | 1348 | 1487 |
| LMArena Korean | 1308 | 1458 |
| LMArena Russian | 1373 | 1482 |
| LMArena Spanish | 1386 | 1472 |
Instruction Following Kimi K3 leads
GLM-4.5-Air: 69.6 (#171), Kimi K3: 77.7 (#14)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1354 | 1483 |
| IFEval | 81.2% | — |
Long Context Kimi K3 leads
GLM-4.5-Air: 41.6 (#135), Kimi K3: 45.8 (#29)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1366 | 1494 |
Writing & Preference Kimi K3 leads
GLM-4.5-Air: 55.9 (#139), Kimi K3: 76.6 (#4)
| Benchmark | GLM-4.5-Air | Kimi K3 |
|---|---|---|
| LMArena Text | 1384 | 1476 |
| LMArena Creative Writing | 1343 | 1454 |
| LMArena Multi-Turn | 1371 | 1488 |
| EQ-Bench Creative Writing | — | 2082 |
| WildBench | 78.9% | — |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is GLM-4.5-Air better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 38.9 on the Noometry Index. GLM-4.5-Air costs 14× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, GLM-4.5-Air or Kimi K3?
GLM-4.5-Air is cheaper. It lists at $0.20 per million input tokens and $1.10 per million output tokens; Kimi K3 lists at $3 and $15.
Is GLM-4.5-Air or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 33.3 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 131K.
How many benchmarks do GLM-4.5-Air and Kimi K3 share?
18 benchmarks have published results for both models. GLM-4.5-Air has 27 scored results on Noometry and Kimi K3 has 53.