Model comparison
Llama2 70b Steerlm Chat vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 31.8 on the Noometry Index.
Last verified . 9 shared benchmarks.
NVIDIA
StepFunSummary
- They share 9 benchmarks with published results for both. Llama2 70b Steerlm Chat scores higher in 0 categories and Step 5 Preview in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Step 5 Preview leads 62.8 to 31.6.
- Llama2 70b Steerlm Chat has downloadable open weights; the other is API-only.
Side by side
| Llama2 70b Steerlm Chat | Step 5 Preview | |
|---|---|---|
| Provider | NVIDIA | StepFun |
| Noometry Index | 31.8 | 47.9 |
| Released | — | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | — | 1.02M |
| Max output | — | 66K |
| Input $ / M tokens | — | $1 |
| Output $ / M tokens | — | $2.70 |
| Results tracked | 9 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 5 Preview leads
Llama2 70b Steerlm Chat: 29.9 (#300), Step 5 Preview: 51.8 (#36)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Coding | 1025 | 1480 |
| LMArena WebDev | — | 1564 |
| SciCode | — | 58.9% |
Reasoning Step 5 Preview leads
Llama2 70b Steerlm Chat: 20.0 (#246), Step 5 Preview: 40.0 (#57)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Hard Prompts | 1047 | 1465 |
| CritPt | — | 20.9% |
Math Step 5 Preview leads
Llama2 70b Steerlm Chat: 31.3 (#226), Step 5 Preview: 46.1 (#72)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Math | 1072 | 1470 |
| ProofBench | — | 42% |
Knowledge Not comparable
Llama2 70b Steerlm Chat: —, Step 5 Preview: 41.2 (#112)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Expert | — | 1470 |
Multimodal Not comparable
Llama2 70b Steerlm Chat: —, Step 5 Preview: 41.0 (#41)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual Step 5 Preview leads
Llama2 70b Steerlm Chat: 28.8 (#270), Step 5 Preview: 53.8 (#54)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1063 | 1432 |
| LMArena Chinese | — | 1519 |
| LMArena Russian | — | 1436 |
| LMArena Spanish | — | 1439 |
Instruction Following Step 5 Preview leads
Llama2 70b Steerlm Chat: 54.2 (#279), Step 5 Preview: 76.0 (#50)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1060 | 1444 |
Long Context Step 5 Preview leads
Llama2 70b Steerlm Chat: 30.4 (#288), Step 5 Preview: 44.5 (#64)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 998 | 1455 |
Writing & Preference Step 5 Preview leads
Llama2 70b Steerlm Chat: 31.6 (#283), Step 5 Preview: 62.8 (#71)
| Benchmark | Llama2 70b Steerlm Chat | Step 5 Preview |
|---|---|---|
| LMArena Text | 1098 | 1442 |
| LMArena Creative Writing | 1091 | 1410 |
| LMArena Multi-Turn | 1058 | 1447 |
Frequently asked questions
Is Llama2 70b Steerlm Chat better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 31.8 on the Noometry Index.
Is Llama2 70b Steerlm Chat or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 29.9 in the Noometry coding category.
How many benchmarks do Llama2 70b Steerlm Chat and Step 5 Preview share?
9 benchmarks have published results for both models. Llama2 70b Steerlm Chat has 9 scored results on Noometry and Step 5 Preview has 18.