Model comparison
DeepSeek LLM 67B vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 24.9 on the Noometry Index.
Last verified . 10 shared benchmarks.
DeepSeek
StepFunSummary
- They share 10 benchmarks with published results for both. DeepSeek LLM 67B scores higher in 0 categories and Step 5 Preview in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Step 5 Preview leads 46.1 to 8.7.
- DeepSeek LLM 67B has downloadable open weights; the other is API-only.
Side by side
| DeepSeek LLM 67B | Step 5 Preview | |
|---|---|---|
| Provider | DeepSeek | StepFun |
| Noometry Index | 24.9 | 47.9 |
| Released | 2023-11-29 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | — | 1.02M |
| Max output | — | 66K |
| Input $ / M tokens | — | $1 |
| Output $ / M tokens | — | $2.70 |
| Results tracked | 15 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 5 Preview leads
DeepSeek LLM 67B: 31.9 (#278), Step 5 Preview: 51.8 (#36)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| LMArena Coding | 1096 | 1480 |
| LMArena WebDev | — | 1564 |
| SciCode | — | 58.9% |
Reasoning Step 5 Preview leads
DeepSeek LLM 67B: 16.5 (#304), Step 5 Preview: 40.0 (#57)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| LMArena Hard Prompts | 1070 | 1465 |
| CritPt | — | 20.9% |
| Chess Puzzles | 0% | — |
| Epoch Capabilities Index | 110.5 | — |
Math Step 5 Preview leads
DeepSeek LLM 67B: 8.7 (#324), Step 5 Preview: 46.1 (#72)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| LMArena Math | 1108 | 1470 |
| OTIS Mock AIME 2024-2025 | 0.8% | — |
| ProofBench | — | 42% |
| MATH Level 5 | 6.4% | — |
Knowledge Step 5 Preview leads
DeepSeek LLM 67B: 7.0 (#313), Step 5 Preview: 41.2 (#112)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| GPQA Diamond | 24.6% | — |
| LMArena Expert | — | 1470 |
Multimodal Not comparable
DeepSeek LLM 67B: —, Step 5 Preview: 41.0 (#41)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual Step 5 Preview leads
DeepSeek LLM 67B: 29.4 (#267), Step 5 Preview: 53.8 (#54)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1073 | 1432 |
| LMArena Chinese | 1132 | 1519 |
| LMArena Russian | — | 1436 |
| LMArena Spanish | — | 1439 |
Instruction Following Step 5 Preview leads
DeepSeek LLM 67B: 55.4 (#277), Step 5 Preview: 76.0 (#50)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1079 | 1444 |
Long Context Step 5 Preview leads
DeepSeek LLM 67B: 33.1 (#265), Step 5 Preview: 44.5 (#64)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1092 | 1455 |
Writing & Preference Step 5 Preview leads
DeepSeek LLM 67B: 31.6 (#282), Step 5 Preview: 62.8 (#71)
| Benchmark | DeepSeek LLM 67B | Step 5 Preview |
|---|---|---|
| LMArena Text | 1105 | 1442 |
| LMArena Creative Writing | 1067 | 1410 |
| LMArena Multi-Turn | 1082 | 1447 |
Frequently asked questions
Is DeepSeek LLM 67B better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 24.9 on the Noometry Index.
Is DeepSeek LLM 67B or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 31.9 in the Noometry coding category.
How many benchmarks do DeepSeek LLM 67B and Step 5 Preview share?
10 benchmarks have published results for both models. DeepSeek LLM 67B has 15 scored results on Noometry and Step 5 Preview has 18.