Model comparison
GPT-5.6 Terra vs Kimi K3
GPT-5.6 Terra and Kimi K3 score almost the same on the Noometry Index (59.2 vs 59.5), so choose on price, context window or the category you care about most.
Last verified . 47 shared benchmarks.
Summary
- They share 47 benchmarks with published results for both. GPT-5.6 Terra scores higher in 2 categories and Kimi K3 in 8 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where GPT-5.6 Terra leads 47.3 to 37.8.
- The biggest single-benchmark swing is FrontierMath Tier 4: 70.7% for GPT-5.6 Terra and 39% for Kimi K3.
- GPT-5.6 Terra is cheaper at $2 / $12 per million input/output tokens, against $3 / $15 for Kimi K3.
- GPT-5.6 Terra accepts more context: 1.05M tokens versus 1.05M.
- Kimi K3 has downloadable open weights; the other is API-only.
Side by side
| GPT-5.6 Terra | Kimi K3 | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 59.2 | 59.5 |
| Released | 2026-07-09 | 2026-07-16 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 1.05M |
| Max output | 128K | 1.05M |
| Input $ / M tokens | $2 | $3 |
| Output $ / M tokens | $12 | $15 |
| Results tracked | 52 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
GPT-5.6 Terra: 57.7 (#19), Kimi K3: 61.0 (#10)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| DeepSWE | 69.6% | 68.5% |
| FrontierCode | 41.3% | 44.2% |
| LMArena WebDev | 1522 | 1654 |
| SciCode | 55% | 59.5% |
| WeirdML | 78.3% | 82.6% |
| LMArena Coding | 1484 | 1508 |
| ALE-Bench | 1,951 | 1,524 |
| CursorBench | 41.3% | — |
| FrontierSWE | — | 25.9% |
Agentic & Tool Use Kimi K3 leads
GPT-5.6 Terra: 40.1 (#25), Kimi K3: 41.8 (#20)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| APEX-Agents | 58.2% | 50.6% |
| GDP.pdf | 24.7% | 19% |
| Vending-Bench 2 | 7,343 | 5,165 |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| BALROG | 53.2% | — |
| GBAEval | — | 48.3% |
Reasoning Kimi K3 leads
GPT-5.6 Terra: 60.7 (#21), Kimi K3: 63.0 (#17)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| ARC-AGI-2 | 83.9% | 60.4% |
| SimpleBench | 48.9% | 60.7% |
| NYT Connections (extended) | 78.4% | 93.6% |
| ARC-AGI-1 | 96.5% | 94.5% |
| CritPt | 30% | 23.4% |
| Chess Puzzles | 54% | 39% |
| LMArena Hard Prompts | 1468 | 1496 |
| Mystery Game Puzzles | 35% | 26% |
| DTBench | 93.3% | 91.2% |
| LMCA | 55% | 52.7% |
| Surface Evolver Bench | 83.8% | 95% |
| Epoch Capabilities Index | 159.62 | 157.45 |
| Kagi LLM Benchmark | 51.3% | — |
| ForecastBench | — | 61.1 |
Math GPT-5.6 Terra leads
GPT-5.6 Terra: 81.6 (#12), Kimi K3: 74.2 (#16)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 86% | 72.2% |
| FrontierMath Tier 4 | 70.7% | 39% |
| OTIS Mock AIME 2024-2025 | 99.7% | 97.2% |
| ProofBench | 74% | 87% |
| LMArena Math | 1466 | 1491 |
| MathArena Final-Answer Competitions | — | 87.8% |
Knowledge Kimi K3 leads
GPT-5.6 Terra: 61.2 (#30), Kimi K3: 63.2 (#21)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| GPQA Diamond | 93.3% | 93.1% |
| SimpleQA Verified | 43.2% | 50.6% |
| LMArena Expert | 1492 | 1521 |
Multimodal GPT-5.6 Terra leads
GPT-5.6 Terra: 47.3 (#11), Kimi K3: 37.8 (#70)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| Blueprint-Bench 2 | 30.8% | 29.5% |
| Furniture Assembly | 54.2% | 34.2% |
| LMArena Vision | 1271 | — |
| LMArena Document | 1472 | — |
Multilingual Kimi K3 leads
GPT-5.6 Terra: 54.4 (#44), Kimi K3: 56.3 (#21)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1439 | 1466 |
| LMArena Chinese | 1513 | 1529 |
| LMArena French | 1471 | 1491 |
| LMArena German | 1460 | 1488 |
| LMArena Japanese | 1457 | 1487 |
| LMArena Korean | 1425 | 1458 |
| LMArena Russian | 1450 | 1482 |
| LMArena Spanish | 1448 | 1472 |
Instruction Following Kimi K3 leads
GPT-5.6 Terra: 76.4 (#40), Kimi K3: 77.7 (#14)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1454 | 1483 |
Long Context Kimi K3 leads
GPT-5.6 Terra: 44.4 (#68), Kimi K3: 45.8 (#29)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1451 | 1494 |
Writing & Preference Kimi K3 leads
GPT-5.6 Terra: 70.2 (#23), Kimi K3: 76.6 (#4)
| Benchmark | GPT-5.6 Terra | Kimi K3 |
|---|---|---|
| LMArena Text | 1447 | 1476 |
| LMArena Creative Writing | 1410 | 1454 |
| EQ-Bench Creative Writing | 1855 | 2082 |
| EQ-Bench 4 | 1234 | 1339 |
| LMArena Multi-Turn | 1449 | 1488 |
Frequently asked questions
Is GPT-5.6 Terra better than Kimi K3?
GPT-5.6 Terra and Kimi K3 score almost the same on the Noometry Index (59.2 vs 59.5), so choose on price, context window or the category you care about most.
Which is cheaper, GPT-5.6 Terra or Kimi K3?
GPT-5.6 Terra is cheaper. It lists at $2 per million input tokens and $12 per million output tokens; Kimi K3 lists at $3 and $15.
Is GPT-5.6 Terra or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 57.7 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Terra does, with 1.05M tokens against 1.05M.
How many benchmarks do GPT-5.6 Terra and Kimi K3 share?
47 benchmarks have published results for both models. GPT-5.6 Terra has 52 scored results on Noometry and Kimi K3 has 53.