Model comparison
Kimi K3 vs Step 5 Preview
Kimi K3 is the stronger model overall, scoring 59.5 to 47.9 on the Noometry Index. Step 5 Preview costs 4.2× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 17 shared benchmarks.
StepFunSummary
- They share 17 benchmarks with published results for both. Kimi K3 scores higher in 8 categories and Step 5 Preview in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.2 to 46.1.
- The biggest single-benchmark swing is ProofBench: 87% for Kimi K3 and 42% for Step 5 Preview.
- Step 5 Preview is cheaper at $1 / $2.70 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 1.02M.
- Kimi K3 has downloadable open weights; the other is API-only.
Side by side
| Kimi K3 | Step 5 Preview | |
|---|---|---|
| Provider | Moonshot AI | StepFun |
| Noometry Index | 59.5 | 47.9 |
| Released | 2026-07-16 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | 1.05M | 1.02M |
| Max output | 1.05M | 66K |
| Input $ / M tokens | $3 | $1 |
| Output $ / M tokens | $15 | $2.70 |
| Results tracked | 53 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Kimi K3: 61.0 (#10), Step 5 Preview: 51.8 (#36)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| LMArena WebDev | 1654 | 1564 |
| SciCode | 59.5% | 58.9% |
| LMArena Coding | 1508 | 1480 |
| DeepSWE | 68.5% | — |
| FrontierCode | 44.2% | — |
| FrontierSWE | 25.9% | — |
| WeirdML | 82.6% | — |
| ALE-Bench | 1,524 | — |
Agentic & Tool Use Not comparable
Kimi K3: 41.8 (#20), Step 5 Preview: —
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| APEX-Agents | 50.6% | — |
| τ²-bench Banking | 37.1% | — |
| PostTrainBench | 32% | — |
| GBAEval | 48.3% | — |
| GDP.pdf | 19% | — |
| Vending-Bench 2 | 5,165 | — |
Reasoning Kimi K3 leads
Kimi K3: 63.0 (#17), Step 5 Preview: 40.0 (#57)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| CritPt | 23.4% | 20.9% |
| LMArena Hard Prompts | 1496 | 1465 |
| ARC-AGI-2 | 60.4% | — |
| SimpleBench | 60.7% | — |
| NYT Connections (extended) | 93.6% | — |
| ARC-AGI-1 | 94.5% | — |
| Chess Puzzles | 39% | — |
| Mystery Game Puzzles | 26% | — |
| DTBench | 91.2% | — |
| LMCA | 52.7% | — |
| Surface Evolver Bench | 95% | — |
| Epoch Capabilities Index | 157.45 | — |
| ForecastBench | 61.1 | — |
Math Kimi K3 leads
Kimi K3: 74.2 (#16), Step 5 Preview: 46.1 (#72)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| ProofBench | 87% | 42% |
| LMArena Math | 1491 | 1470 |
| FrontierMath (Tiers 1-3) | 72.2% | — |
| FrontierMath Tier 4 | 39% | — |
| MathArena Final-Answer Competitions | 87.8% | — |
| OTIS Mock AIME 2024-2025 | 97.2% | — |
Knowledge Kimi K3 leads
Kimi K3: 63.2 (#21), Step 5 Preview: 41.2 (#112)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1521 | 1470 |
| GPQA Diamond | 93.1% | — |
| SimpleQA Verified | 50.6% | — |
Multimodal Step 5 Preview leads
Kimi K3: 37.8 (#70), Step 5 Preview: 41.0 (#41)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
| Blueprint-Bench 2 | 29.5% | — |
| Furniture Assembly | 34.2% | — |
Multilingual Kimi K3 leads
Kimi K3: 56.3 (#21), Step 5 Preview: 53.8 (#54)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1466 | 1432 |
| LMArena Chinese | 1529 | 1519 |
| LMArena Russian | 1482 | 1436 |
| LMArena Spanish | 1472 | 1439 |
| LMArena French | 1491 | — |
| LMArena German | 1488 | — |
| LMArena Japanese | 1487 | — |
| LMArena Korean | 1458 | — |
Instruction Following Kimi K3 leads
Kimi K3: 77.7 (#14), Step 5 Preview: 76.0 (#50)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1483 | 1444 |
Long Context Kimi K3 leads
Kimi K3: 45.8 (#29), Step 5 Preview: 44.5 (#64)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1494 | 1455 |
Writing & Preference Kimi K3 leads
Kimi K3: 76.6 (#4), Step 5 Preview: 62.8 (#71)
| Benchmark | Kimi K3 | Step 5 Preview |
|---|---|---|
| LMArena Text | 1476 | 1442 |
| LMArena Creative Writing | 1454 | 1410 |
| LMArena Multi-Turn | 1488 | 1447 |
| EQ-Bench Creative Writing | 2082 | — |
| EQ-Bench 4 | 1339 | — |
Frequently asked questions
Is Kimi K3 better than Step 5 Preview?
Kimi K3 is the stronger model overall, scoring 59.5 to 47.9 on the Noometry Index. Step 5 Preview costs 4.2× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, Kimi K3 or Step 5 Preview?
Step 5 Preview is cheaper. It lists at $1 per million input tokens and $2.70 per million output tokens; Kimi K3 lists at $3 and $15.
Is Kimi K3 or Step 5 Preview better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 51.8 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 1.02M.
How many benchmarks do Kimi K3 and Step 5 Preview share?
17 benchmarks have published results for both models. Kimi K3 has 53 scored results on Noometry and Step 5 Preview has 18.