# GPT-5.6 Sol vs Kimi K2.5

> GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 47.8 on the Noometry Index. Kimi K2.5 costs 8.9× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-6-sol-vs-kimi-k2-5
- Last updated: 2026-10-10
- Shared benchmarks: 38

## Summary

- They share 38 benchmarks with published results for both. GPT-5.6 Sol scores higher in 9 categories and Kimi K2.5 in 1 category; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.6 Sol leads 74.7 to 31.2.
- The biggest single-benchmark swing is ARC-AGI-2: 92.5% for GPT-5.6 Sol and 11.8% for Kimi K2.5.
- Kimi K2.5 is cheaper at $0.45 / $2.25 per million input/output tokens, against $4 / $20 for GPT-5.6 Sol.
- GPT-5.6 Sol accepts more context: 1.05M tokens versus 262K.
- Kimi K2.5 has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 64.4 | 47.8 |
| Rank | 6 | 50 |
| Context | 1.05M | 262K |
| Input $/M | $4 | $0.45 |
| Output $/M | $20 | $2.25 |
| Weights | Proprietary | Open |

## Coding

- GPT-5.6 Sol: 61.1 (#7)
- Kimi K2.5: 48.7 (#48)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena WebDev | 1619 | 1436 |
| SciCode | 57.1% | 49% |
| WeirdML | 89.4% | 45.6% |
| LMArena Coding | 1493 | 1474 |
| ALE-Bench | 2,177 | 821.65 |
| SWE-bench Verified | — | 73.8% |
| DeepSWE | 72.7% | — |
| FrontierCode | 47.5% | — |
| SWE-bench Verified (bash only) | — | 70.8% |
| CursorBench | 41.7% | — |
| SWE-bench Multilingual | — | 67.3% |
| FrontierSWE | 32.2% | — |
| MirrorCode | 20% | — |

## Agentic & Tool Use

- GPT-5.6 Sol: 50.2 (#7)
- Kimi K2.5: 34.0 (#44)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| Vending-Bench 2 | 9,619 | 1,198 |
| Terminal-Bench | — | 43.2% |
| APEX-Agents | 51.4% | — |
| OSWorld 2.0 | 27.3% | — |
| τ²-bench Banking | 46.9% | — |
| OSWorld | — | 63.3% |
| PostTrainBench | 36.2% | — |
| BALROG | 60% | — |
| GBAEval | 52.6% | — |
| GDP.pdf | 30.7% | — |
| LMArena Search | 1257 | — |

## Reasoning

- GPT-5.6 Sol: 74.7 (#6)
- Kimi K2.5: 31.2 (#74)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| ARC-AGI-2 | 92.5% | 11.8% |
| SimpleBench | 71.7% | 46.8% |
| Kagi LLM Benchmark | 67% | 78.5% |
| NYT Connections (extended) | 93.8% | 69.9% |
| ARC-AGI-1 | 97.5% | 65.3% |
| CritPt | 32.3% | 3.1% |
| Chess Puzzles | 64% | 12% |
| EnigmaEval | 37.1% | 3.4% |
| LMArena Hard Prompts | 1481 | 1454 |
| Epoch Capabilities Index | 161.99 | 148.05 |
| Thematic Generalization | — | 69.4% |
| EBR-Bench | 44.8% | — |
| Mystery Game Puzzles | 58% | — |
| DTBench | 96% | — |
| LMCA | 59.2% | — |
| Surface Evolver Bench | 93.1% | — |
| Bench to the Future 3 | 0.14 | — |

## Math

- GPT-5.6 Sol: 85.5 (#7)
- Kimi K2.5: 50.3 (#46)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 92.2% |
| LMArena Math | 1472 | 1470 |
| FrontierMath (Tiers 1-3) | 89.1% | — |
| FrontierMath Tier 4 | 82.9% | — |
| MathArena Final-Answer Competitions | — | 62.3% |
| ProofBench | 83% | — |
| FrontierMath (Feb 2025 set) | — | 27.9% |
| FrontierMath Erdős | 0% | — |
| FrontierMath Tier 4 (v1) | — | 4.2% |

## Knowledge

- GPT-5.6 Sol: 64.3 (#16)
- Kimi K2.5: 53.6 (#52)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| GPQA Diamond | 93.5% | 87.6% |
| SimpleQA Verified | 69.7% | 34.3% |
| Vectara Hallucination Rate | 12.4% | 14.2% |
| LMArena Expert | 1517 | 1466 |
| Humanity's Last Exam | — | 24.4% |

## Multimodal

- GPT-5.6 Sol: 48.9 (#8)
- Kimi K2.5: 40.4 (#37)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Vision | 1281 | 1269 |
| LMArena Document | 1483 | 1430 |
| Blueprint-Bench 2 | 33.6% | — |
| Furniture Assembly | 56.7% | — |

## Multilingual

- GPT-5.6 Sol: 55.4 (#32)
- Kimi K2.5: 54.1 (#50)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Non-English | 1450 | 1433 |
| LMArena Chinese | 1523 | 1495 |
| LMArena French | 1475 | 1454 |
| LMArena German | 1476 | 1441 |
| LMArena Japanese | 1475 | 1422 |
| LMArena Korean | 1434 | 1412 |
| LMArena Russian | 1466 | 1436 |
| LMArena Spanish | 1449 | 1450 |

## Instruction Following

- GPT-5.6 Sol: 77.8 (#16)
- Kimi K2.5: 75.5 (#60)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Instruction Following | 1480 | 1431 |

## Long Context

- GPT-5.6 Sol: 45.3 (#42)
- Kimi K2.5: 52.1 (#7)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Longer Query | 1479 | 1445 |
| Fiction.LiveBench | — | 86.1% |
| CL-bench | — | 19.3% |
| CL-bench Life | — | 13.2% |

## Writing & Preference

- GPT-5.6 Sol: 73.5 (#12)
- Kimi K2.5: 65.3 (#51)

| Benchmark | GPT-5.6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Text | 1456 | 1445 |
| LMArena Creative Writing | 1450 | 1423 |
| EQ-Bench Creative Writing | 1972 | 1579 |
| LMArena Multi-Turn | 1456 | 1444 |
| EQ-Bench 4 | 1250 | — |

## FAQ

### Is GPT-5.6 Sol better than Kimi K2.5?

GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 47.8 on the Noometry Index. Kimi K2.5 costs 8.9× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.6 Sol 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; GPT-5.6 Sol lists at $4 and $20.

### Is GPT-5.6 Sol or Kimi K2.5 better for coding?

GPT-5.6 Sol scores higher on coding benchmarks: 61.1 versus 48.7 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.6 Sol does, with 1.05M tokens against 262K.

### How many benchmarks do GPT-5.6 Sol and Kimi K2.5 share?

38 benchmarks have published results for both models. GPT-5.6 Sol has 64 scored results on Noometry and Kimi K2.5 has 51.