Model comparison
DeepSeek-R1-Distill-Llama-70B vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 37.8 on the Noometry Index.
Last verified . 1 shared benchmarks.
DeepSeek
StepFunSummary
- They share 1 benchmark with published results for both. DeepSeek-R1-Distill-Llama-70B scores higher in 0 categories and Step 3 in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Step 3 leads 36.8 to 30.7.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 52.3% for DeepSeek-R1-Distill-Llama-70B and 62.3% for Step 3.
Side by side
| DeepSeek-R1-Distill-Llama-70B | Step 3 | |
|---|---|---|
| Provider | DeepSeek | StepFun |
| Noometry Index | 37.8 | 40.5 |
| Released | 2025-01-20 | — |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 13 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
DeepSeek-R1-Distill-Llama-70B: 36.8 (#202), Step 3: 40.1 (#147)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| BigCodeBench Instruct | 35.3% | — |
| LiveBench Coding | 51.6% | — |
| LMArena Coding | — | 1367 |
| BigCodeBench Complete | 49.9% | — |
Reasoning Step 3 leads
DeepSeek-R1-Distill-Llama-70B: 24.9 (#156), Step 3: 28.4 (#105)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | 52.3% | 62.3% |
| LiveBench Reasoning | 67.6% | — |
| LMArena Hard Prompts | — | 1355 |
| LiveBench Data Analysis | 55.9% | — |
| LiveBench | 54.5% | — |
Math Step 3 leads
DeepSeek-R1-Distill-Llama-70B: 36.0 (#176), Step 3: 37.6 (#148)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 51.4% | — |
| LiveBench Math | 58.1% | — |
| LMArena Math | — | 1366 |
| MATH Level 5 | 89.9% | — |
Knowledge Step 3 leads
DeepSeek-R1-Distill-Llama-70B: 30.7 (#225), Step 3: 36.8 (#164)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| GPQA Diamond | 55.7% | — |
| LMArena Expert | — | 1333 |
Multimodal Not comparable
DeepSeek-R1-Distill-Llama-70B: —, Step 3: 35.5 (#86)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual Not comparable
DeepSeek-R1-Distill-Llama-70B: —, Step 3: 46.3 (#159)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| LMArena Non-English | — | 1327 |
| LMArena Chinese | — | 1397 |
| LMArena German | — | 1371 |
| LMArena Korean | — | 1269 |
| LMArena Russian | — | 1331 |
| LMArena Spanish | — | 1371 |
Instruction Following Step 3 leads
DeepSeek-R1-Distill-Llama-70B: 68.2 (#190), Step 3: 70.4 (#164)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| LiveBench Instruction Following | 69.9% | — |
| LMArena Instruction Following | — | 1332 |
Long Context Not comparable
DeepSeek-R1-Distill-Llama-70B: —, Step 3: 40.3 (#157)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| LMArena Longer Query | — | 1326 |
Writing & Preference Step 3 leads
DeepSeek-R1-Distill-Llama-70B: 49.0 (#194), Step 3: 54.3 (#151)
| Benchmark | DeepSeek-R1-Distill-Llama-70B | Step 3 |
|---|---|---|
| LMArena Text | — | 1350 |
| LMArena Creative Writing | — | 1321 |
| LMArena Multi-Turn | — | 1341 |
| LiveBench Language | 23.8% | — |
Frequently asked questions
Is DeepSeek-R1-Distill-Llama-70B better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 37.8 on the Noometry Index.
Is DeepSeek-R1-Distill-Llama-70B or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 36.8 in the Noometry coding category.
How many benchmarks do DeepSeek-R1-Distill-Llama-70B and Step 3 share?
1 benchmark has published results for both models. DeepSeek-R1-Distill-Llama-70B has 13 scored results on Noometry and Step 3 has 17.