Model comparison
GPT-5.6 Terra vs Kimi K2.6
GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 47.5 on the Noometry Index. Kimi K2.6 costs 2.6× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.
Last verified . 42 shared benchmarks.
Summary
- They share 42 benchmarks with published results for both. GPT-5.6 Terra scores higher in 8 categories and Kimi K2.6 in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Terra leads 81.5 to 56.8.
- The biggest single-benchmark swing is ProofBench: 74% for GPT-5.6 Terra and 16% for Kimi K2.6.
- Kimi K2.6 is cheaper at $0.95 / $4 per million input/output tokens, against $2 / $12 for GPT-5.6 Terra.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 262K.
- Kimi K2.6 has downloadable open weights; the other is API-only.
Side by side
| GPT-5.6 Terra | Kimi K2.6 | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 59.1 | 47.5 |
| Released | 2026-07-09 | 2026-04-20 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 128K | 262K |
| Input $ / M tokens | $2 | $0.95 |
| Output $ / M tokens | $12 | $4 |
| Results tracked | 52 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Terra leads
GPT-5.6 Terra: 57.1 (#16), Kimi K2.6: 50.5 (#41)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| LMArena WebDev | 1520 | 1509 |
| SciCode | 55% | 53.5% |
| WeirdML | 78.3% | 55.9% |
| LMArena Coding | 1484 | 1488 |
| ALE-Bench | 1,951 | 1,093 |
| SWE-bench Verified | — | 76.7% |
| DeepSWE | 69.6% | — |
| FrontierCode | 41.3% | — |
| CursorBench | 41.3% | — |
Agentic & Tool Use GPT-5.6 Terra leads
GPT-5.6 Terra: 40.2 (#23), Kimi K2.6: 21.8 (#128)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| GDP.pdf | 24.7% | 12% |
| Vending-Bench 2 | 7,343 | 6,205 |
| APEX-Agents | 58.2% | — |
| OSWorld 2.0 | — | 4.6% |
| BALROG | 53.2% | — |
| ExploitBench | — | 18.4% |
| GBAEval | — | 0.9% |
Reasoning GPT-5.6 Terra leads
GPT-5.6 Terra: 60.6 (#19), Kimi K2.6: 39.6 (#51)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| NYT Connections (extended) | 78.4% | 87.2% |
| CritPt | 30% | 8% |
| Chess Puzzles | 54% | 26% |
| LMArena Hard Prompts | 1467 | 1470 |
| Mystery Game Puzzles | 35% | 18% |
| DTBench | 93.3% | 90.9% |
| LMCA | 55% | 37.3% |
| Epoch Capabilities Index | 159.31 | 151.03 |
| ARC-AGI-2 | 83.9% | — |
| SimpleBench | 48.9% | — |
| Kagi LLM Benchmark | 51.3% | — |
| ARC-AGI-1 | 96.5% | — |
| EBR-Bench | — | 2.4% |
| Surface Evolver Bench | 83.8% | — |
Math GPT-5.6 Terra leads
GPT-5.6 Terra: 81.5 (#10), Kimi K2.6: 56.8 (#34)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 86% | 57.2% |
| FrontierMath Tier 4 | 70.7% | 25.6% |
| OTIS Mock AIME 2024-2025 | 99.7% | 96.1% |
| ProofBench | 74% | 16% |
| LMArena Math | 1463 | 1475 |
| MathArena Final-Answer Competitions | — | 72.9% |
| FrontierMath (Feb 2025 set) | — | 39% |
| FrontierMath Tier 4 (v1) | — | 14.6% |
Knowledge GPT-5.6 Terra leads
GPT-5.6 Terra: 61.3 (#27), Kimi K2.6: 53.9 (#48)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| GPQA Diamond | 93.3% | 90.8% |
| SimpleQA Verified | 43.2% | 34.9% |
| LMArena Expert | 1492 | 1491 |
| Vectara Hallucination Rate | — | 10.8% |
Multimodal GPT-5.6 Terra leads
GPT-5.6 Terra: 47.5 (#10), Kimi K2.6: 31.8 (#95)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| LMArena Vision | 1271 | 1282 |
| Blueprint-Bench 2 | 30.8% | 3.9% |
| Furniture Assembly | 54.2% | 21.7% |
| LMArena Document | 1472 | 1451 |
Multilingual Too close to call
GPT-5.6 Terra: 54.5 (#42), Kimi K2.6: 55.1 (#35)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| LMArena Non-English | 1439 | 1446 |
| LMArena Chinese | 1514 | 1521 |
| LMArena French | 1476 | 1471 |
| LMArena German | 1462 | 1449 |
| LMArena Japanese | 1446 | 1442 |
| LMArena Korean | 1420 | 1428 |
| LMArena Russian | 1450 | 1446 |
| LMArena Spanish | 1455 | 1463 |
Instruction Following Too close to call
GPT-5.6 Terra: 76.6 (#38), Kimi K2.6: 76.5 (#40)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| LMArena Instruction Following | 1453 | 1451 |
Long Context Too close to call
GPT-5.6 Terra: 44.4 (#64), Kimi K2.6: 44.9 (#50)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| LMArena Longer Query | 1452 | 1468 |
Writing & Preference GPT-5.6 Terra leads
GPT-5.6 Terra: 70.4 (#21), Kimi K2.6: 68.7 (#25)
| Benchmark | GPT-5.6 Terra | Kimi K2.6 |
|---|---|---|
| LMArena Text | 1446 | 1455 |
| LMArena Creative Writing | 1411 | 1434 |
| EQ-Bench Creative Writing | 1855 | 1725 |
| EQ-Bench 4 | 1234 | 1202 |
| LMArena Multi-Turn | 1451 | 1453 |
Frequently asked questions
Is GPT-5.6 Terra better than Kimi K2.6?
GPT-5.6 Terra is the stronger model overall, scoring 59.1 to 47.5 on the Noometry Index. Kimi K2.6 costs 2.6× less per token, which makes it the better buy when GPT-5.6 Terra's lead doesn't matter for your workload.
Which is cheaper, GPT-5.6 Terra or Kimi K2.6?
Kimi K2.6 is cheaper. It lists at $0.95 per million input tokens and $4 per million output tokens; GPT-5.6 Terra lists at $2 and $12.
Is GPT-5.6 Terra or Kimi K2.6 better for coding?
GPT-5.6 Terra scores higher on coding benchmarks: 57.1 versus 50.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Terra does, with 1.05M tokens against 262K.
How many benchmarks do GPT-5.6 Terra and Kimi K2.6 share?
42 benchmarks have published results for both models. GPT-5.6 Terra has 52 scored results on Noometry and Kimi K2.6 has 51.