Model comparison
GLM-5 vs Kimi K2.5
Kimi K2.5 is the stronger model overall, scoring 48.1 to 46.1 on the Noometry Index.
Last verified . 40 shared benchmarks.
Summary
- They share 40 benchmarks with published results for both. GLM-5 scores higher in 2 categories and Kimi K2.5 in 7 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in long context, where Kimi K2.5 leads 52.1 to 44.7.
- The biggest single-benchmark swing is ARC-AGI-1: 44.7% for GLM-5 and 65.3% for Kimi K2.5.
- Kimi K2.5 is cheaper at $0.45 / $2.25 per million input/output tokens, against $1 / $3.20 for GLM-5.
- Kimi K2.5 accepts more context: 262K tokens versus 205K.
Side by side
| GLM-5 | Kimi K2.5 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Moonshot AI |
| Noometry Index | 46.1 | 48.1 |
| Released | 2026-02-11 | 2026-01-27 |
| Weights | Open | Open |
| Context window | 205K | 262K |
| Max output | 131K | 262K |
| Input $ / M tokens | $1 | $0.45 |
| Output $ / M tokens | $3.20 | $2.25 |
| Results tracked | 45 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GLM-5: 49.0 (#52), Kimi K2.5: 48.8 (#53)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| SWE-bench Verified | 72.1% | 73.8% |
| SWE-bench Verified (bash only) | 72.8% | 70.8% |
| LMArena WebDev | 1434 | 1437 |
| SWE-bench Multilingual | 69.7% | 67.3% |
| WeirdML | 48.2% | 45.6% |
| LMArena Coding | 1461 | 1474 |
| ALE-Bench | 765.62 | 821.65 |
| SciCode | — | 49% |
Agentic & Tool Use Kimi K2.5 leads
GLM-5: 31.1 (#71), Kimi K2.5: 34.2 (#48)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| Terminal-Bench | 52.4% | 43.2% |
| Vending-Bench 2 | 4,432 | 1,198 |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Banking | 9.8% | — |
| τ²-bench Retail | 73.7% | — |
| τ²-bench Telecom | 86.8% | — |
| OSWorld | — | 63.3% |
Reasoning Kimi K2.5 leads
GLM-5: 27.6 (#116), Kimi K2.5: 31.2 (#80)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| ARC-AGI-2 | 4.9% | 11.8% |
| SimpleBench | 53.2% | 46.8% |
| Kagi LLM Benchmark | 75% | 78.5% |
| NYT Connections (extended) | 74.8% | 69.9% |
| ARC-AGI-1 | 44.7% | 65.3% |
| Chess Puzzles | 10% | 12% |
| LMArena Hard Prompts | 1452 | 1453 |
| Epoch Capabilities Index | 145.83 | 148.03 |
| CritPt | — | 3.1% |
| EnigmaEval | — | 3.4% |
| Thematic Generalization | — | 69.4% |
| ForecastBench | 61 | — |
Math Kimi K2.5 leads
GLM-5: 46.4 (#71), Kimi K2.5: 51.8 (#53)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| MathArena Final-Answer Competitions | 65.7% | 62.3% |
| OTIS Mock AIME 2024-2025 | 80% | 92.2% |
| LMArena Math | 1440 | 1470 |
| FrontierMath (Feb 2025 set) | 16.4% | 27.9% |
| FrontierMath Tier 4 (v1) | 2.1% | 4.2% |
Knowledge Kimi K2.5 leads
GLM-5: 52.3 (#64), Kimi K2.5: 53.6 (#56)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| GPQA Diamond | 87.8% | 87.6% |
| Vectara Hallucination Rate | 10.1% | 14.2% |
| LMArena Expert | 1454 | 1466 |
| Humanity's Last Exam | — | 24.4% |
| SimpleQA Verified | — | 34.3% |
Multimodal Not comparable
GLM-5: —, Kimi K2.5: 41.1 (#39)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| LMArena Vision | — | 1269 |
| LMArena Document | — | 1430 |
Multilingual Too close to call
GLM-5: 53.7 (#58), Kimi K2.5: 53.9 (#53)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| LMArena Non-English | 1430 | 1433 |
| LMArena Chinese | 1511 | 1495 |
| LMArena French | 1455 | 1454 |
| LMArena German | 1445 | 1441 |
| LMArena Japanese | 1416 | 1421 |
| LMArena Korean | 1423 | 1410 |
| LMArena Russian | 1436 | 1435 |
| LMArena Spanish | 1454 | 1450 |
Instruction Following Too close to call
GLM-5: 75.2 (#67), Kimi K2.5: 75.3 (#64)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| LMArena Instruction Following | 1428 | 1431 |
Long Context Kimi K2.5 leads
GLM-5: 44.7 (#60), Kimi K2.5: 52.1 (#7)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| CL-bench | 18.7% | 19.3% |
| LMArena Longer Query | 1446 | 1445 |
| Fiction.LiveBench | — | 86.1% |
| CL-bench Life | — | 13.2% |
Writing & Preference Too close to call
GLM-5: 66.0 (#38), Kimi K2.5: 65.1 (#53)
| Benchmark | GLM-5 | Kimi K2.5 |
|---|---|---|
| LMArena Text | 1446 | 1445 |
| LMArena Creative Writing | 1439 | 1423 |
| EQ-Bench Creative Writing | 1601 | 1579 |
| LMArena Multi-Turn | 1456 | 1444 |
Frequently asked questions
Is GLM-5 better than Kimi K2.5?
Kimi K2.5 is the stronger model overall, scoring 48.1 to 46.1 on the Noometry Index.
Which is cheaper, GLM-5 or Kimi K2.5?
Kimi K2.5 is cheaper. It lists at $0.45 per million input tokens and $2.25 per million output tokens; GLM-5 lists at $1 and $3.20.
Is GLM-5 or Kimi K2.5 better for coding?
They score almost the same on coding (49.0 vs 48.8); test both on your own repository before choosing.
Which has the bigger context window?
Kimi K2.5 does, with 262K tokens against 205K.
How many benchmarks do GLM-5 and Kimi K2.5 share?
40 benchmarks have published results for both models. GLM-5 has 45 scored results on Noometry and Kimi K2.5 has 51.