Model comparison
GLM-5 vs Kimi K2 (Jul 2025)
GLM-5 is the stronger model overall, scoring 46.1 to 41.2 on the Noometry Index. Kimi K2 (Jul 2025) costs 1.5× less per token, which makes it the better buy when GLM-5's lead doesn't matter for your workload.
Last verified . 30 shared benchmarks.
Summary
- They share 30 benchmarks with published results for both. GLM-5 scores higher in 8 categories and Kimi K2 (Jul 2025) in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5 leads 52.3 to 37.3.
- The biggest single-benchmark swing is SimpleBench: 53.2% for GLM-5 and 26.3% for Kimi K2 (Jul 2025).
- Kimi K2 (Jul 2025) is cheaper at $0.57 / $2.30 per million input/output tokens, against $1 / $3.20 for GLM-5.
- Kimi K2 (Jul 2025) accepts more context: 262K tokens versus 205K.
Side by side
| GLM-5 | Kimi K2 (Jul 2025) | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Moonshot AI |
| Noometry Index | 46.1 | 41.2 |
| Released | 2026-02-11 | 2025-07-12 |
| Weights | Open | Open |
| Context window | 205K | 262K |
| Max output | 131K | 262K |
| Input $ / M tokens | $1 | $0.57 |
| Output $ / M tokens | $3.20 | $2.30 |
| Results tracked | 45 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5 leads
GLM-5: 49.0 (#52), Kimi K2 (Jul 2025): 42.4 (#102)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| SWE-bench Verified (bash only) | 72.8% | 63.4% |
| WeirdML | 48.2% | 42.8% |
| LMArena Coding | 1461 | 1399 |
| ALE-Bench | 765.62 | 597.5 |
| SWE-bench Verified | 72.1% | — |
| Aider Polyglot | — | 59.1% |
| LMArena WebDev | 1434 | — |
| SWE-bench Multilingual | 69.7% | — |
| GSO | — | 4.9% |
Agentic & Tool Use Kimi K2 (Jul 2025) leads
GLM-5: 31.1 (#71), Kimi K2 (Jul 2025): 32.4 (#64)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| Terminal-Bench | 52.4% | 35.7% |
| Berkeley Function Calling Leaderboard | — | 59.1% |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Banking | 9.8% | — |
| τ²-bench Retail | 73.7% | — |
| τ²-bench Telecom | 86.8% | — |
| METR Time Horizons | — | 59.2% |
| Vending-Bench 2 | 4,432 | — |
Reasoning GLM-5 leads
GLM-5: 27.6 (#116), Kimi K2 (Jul 2025): 23.3 (#179)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| SimpleBench | 53.2% | 26.3% |
| Kagi LLM Benchmark | 75% | 64.4% |
| LMArena Hard Prompts | 1452 | 1384 |
| Epoch Capabilities Index | 145.83 | 146.01 |
| ForecastBench | 61 | 60.2 |
| ARC-AGI-2 | 4.9% | — |
| NYT Connections (extended) | 74.8% | — |
| ARC-AGI-1 | 44.7% | — |
| Chess Puzzles | 10% | — |
Math GLM-5 leads
GLM-5: 46.4 (#71), Kimi K2 (Jul 2025): 42.7 (#83)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| LMArena Math | 1440 | 1397 |
| FrontierMath (Feb 2025 set) | 16.4% | 21.4% |
| FrontierMath Tier 4 (v1) | 2.1% | 0% |
| MathArena Final-Answer Competitions | 65.7% | — |
| OTIS Mock AIME 2024-2025 | 80% | — |
| Omni-MATH | — | 65.4% |
Knowledge GLM-5 leads
GLM-5: 52.3 (#64), Kimi K2 (Jul 2025): 37.3 (#157)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| Vectara Hallucination Rate | 10.1% | 17.9% |
| LMArena Expert | 1454 | 1365 |
| GPQA Diamond | 87.8% | — |
| MMLU-Pro | — | 81.9% |
| Confabulations | — | 20.4% |
| GPQA (HELM) | — | 65.3% |
Multilingual GLM-5 leads
GLM-5: 53.7 (#58), Kimi K2 (Jul 2025): 49.6 (#130)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| LMArena Non-English | 1430 | 1372 |
| LMArena Chinese | 1511 | 1415 |
| LMArena French | 1455 | 1379 |
| LMArena German | 1445 | 1387 |
| LMArena Japanese | 1416 | 1349 |
| LMArena Korean | 1423 | 1325 |
| LMArena Russian | 1436 | 1385 |
| LMArena Spanish | 1454 | 1386 |
Instruction Following GLM-5 leads
GLM-5: 75.2 (#67), Kimi K2 (Jul 2025): 71.1 (#156)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| LMArena Instruction Following | 1428 | 1348 |
| IFEval | — | 85% |
Long Context GLM-5 leads
GLM-5: 44.7 (#60), Kimi K2 (Jul 2025): 41.2 (#145)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| CL-bench | 18.7% | 17.6% |
| LMArena Longer Query | 1446 | 1353 |
| Fiction.LiveBench | — | 66.7% |
Writing & Preference GLM-5 leads
GLM-5: 66.0 (#38), Kimi K2 (Jul 2025): 62.3 (#78)
| Benchmark | GLM-5 | Kimi K2 (Jul 2025) |
|---|---|---|
| LMArena Text | 1446 | 1380 |
| LMArena Creative Writing | 1439 | 1350 |
| EQ-Bench Creative Writing | 1601 | 1666 |
| LMArena Multi-Turn | 1456 | 1371 |
| Short-Story Creative Writing | — | 85.6% |
| WildBench | — | 86.2% |
Frequently asked questions
Is GLM-5 better than Kimi K2 (Jul 2025)?
GLM-5 is the stronger model overall, scoring 46.1 to 41.2 on the Noometry Index. Kimi K2 (Jul 2025) costs 1.5× 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 Kimi K2 (Jul 2025)?
Kimi K2 (Jul 2025) is cheaper. It lists at $0.57 per million input tokens and $2.30 per million output tokens; GLM-5 lists at $1 and $3.20.
Is GLM-5 or Kimi K2 (Jul 2025) better for coding?
GLM-5 scores higher on coding benchmarks: 49.0 versus 42.4 in the Noometry coding category.
Which has the bigger context window?
Kimi K2 (Jul 2025) does, with 262K tokens against 205K.
How many benchmarks do GLM-5 and Kimi K2 (Jul 2025) share?
30 benchmarks have published results for both models. GLM-5 has 45 scored results on Noometry and Kimi K2 (Jul 2025) has 42.