Model comparison
Kimi K2.7 Code vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 43.3 on the Noometry Index.
Last verified . 3 shared benchmarks.
StepFunSummary
- They share 3 benchmarks with published results for both. Kimi K2.7 Code scores higher in 2 categories and Step 5 Preview in 2 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Kimi K2.7 Code leads 53.5 to 41.2.
- The biggest single-benchmark swing is SciCode: 47.5% for Kimi K2.7 Code and 58.9% for Step 5 Preview.
- Step 5 Preview is cheaper at $1 / $2.70 per million input/output tokens, against $0.95 / $4 for Kimi K2.7 Code.
- Step 5 Preview accepts more context: 1.02M tokens versus 262K.
- Kimi K2.7 Code has downloadable open weights; the other is API-only.
Side by side
| Kimi K2.7 Code | Step 5 Preview | |
|---|---|---|
| Provider | Moonshot AI | StepFun |
| Noometry Index | 43.3 | 47.9 |
| Released | 2026-06-12 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | 262K | 1.02M |
| Max output | 262K | 66K |
| Input $ / M tokens | $0.95 | $1 |
| Output $ / M tokens | $4 | $2.70 |
| Results tracked | 19 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 5 Preview leads
Kimi K2.7 Code: 42.9 (#95), Step 5 Preview: 51.8 (#36)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| LMArena WebDev | 1473 | 1564 |
| SciCode | 47.5% | 58.9% |
| DeepSWE | 30.5% | — |
| FrontierCode | 30.1% | — |
| WeirdML | 54.1% | — |
| LMArena Coding | — | 1480 |
| ALE-Bench | 886.23 | — |
Agentic & Tool Use Not comparable
Kimi K2.7 Code: 24.0 (#122), Step 5 Preview: —
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| APEX-Agents | 37.6% | — |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 5,083 | — |
Reasoning Step 5 Preview leads
Kimi K2.7 Code: 39.0 (#61), Step 5 Preview: 40.0 (#57)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| CritPt | 10% | 20.9% |
| SimpleBench | 57.9% | — |
| Chess Puzzles | 21% | — |
| LMArena Hard Prompts | — | 1465 |
| Surface Evolver Bench | 48.8% | — |
| Epoch Capabilities Index | 149.97 | — |
Math Kimi K2.7 Code leads
Kimi K2.7 Code: 52.9 (#48), Step 5 Preview: 46.1 (#72)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| FrontierMath (Tiers 1-3) | 54% | — |
| FrontierMath Tier 4 | 12.2% | — |
| OTIS Mock AIME 2024-2025 | 95.6% | — |
| ProofBench | — | 42% |
| LMArena Math | — | 1470 |
Knowledge Kimi K2.7 Code leads
Kimi K2.7 Code: 53.5 (#57), Step 5 Preview: 41.2 (#112)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| GPQA Diamond | 87.9% | — |
| SimpleQA Verified | 36.5% | — |
| LMArena Expert | — | 1470 |
Multimodal Not comparable
Kimi K2.7 Code: —, Step 5 Preview: 41.0 (#41)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual Not comparable
Kimi K2.7 Code: —, Step 5 Preview: 53.8 (#54)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| LMArena Non-English | — | 1432 |
| LMArena Chinese | — | 1519 |
| LMArena Russian | — | 1436 |
| LMArena Spanish | — | 1439 |
Instruction Following Not comparable
Kimi K2.7 Code: —, Step 5 Preview: 76.0 (#50)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | — | 1444 |
Long Context Not comparable
Kimi K2.7 Code: —, Step 5 Preview: 44.5 (#64)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | — | 1455 |
Writing & Preference Not comparable
Kimi K2.7 Code: —, Step 5 Preview: 62.8 (#71)
| Benchmark | Kimi K2.7 Code | Step 5 Preview |
|---|---|---|
| LMArena Text | — | 1442 |
| LMArena Creative Writing | — | 1410 |
| LMArena Multi-Turn | — | 1447 |
Frequently asked questions
Is Kimi K2.7 Code better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 43.3 on the Noometry Index.
Which is cheaper, Kimi K2.7 Code 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 K2.7 Code lists at $0.95 and $4.
Is Kimi K2.7 Code or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 42.9 in the Noometry coding category.
Which has the bigger context window?
Step 5 Preview does, with 1.02M tokens against 262K.
How many benchmarks do Kimi K2.7 Code and Step 5 Preview share?
3 benchmarks have published results for both models. Kimi K2.7 Code has 19 scored results on Noometry and Step 5 Preview has 18.