Model comparison
DeepSeek LLM 67B vs Solar Pro4
Solar Pro4 is the stronger model overall, scoring 42.1 to 24.9 on the Noometry Index.
Last verified . 10 shared benchmarks.
DeepSeek
UpstageSummary
- They share 10 benchmarks with published results for both. DeepSeek LLM 67B scores higher in 0 categories and Solar Pro4 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Solar Pro4 leads 39.8 to 7.0.
- DeepSeek LLM 67B has downloadable open weights; the other is API-only.
Side by side
| DeepSeek LLM 67B | Solar Pro4 | |
|---|---|---|
| Provider | DeepSeek | Upstage |
| Noometry Index | 24.9 | 42.1 |
| Released | 2023-11-29 | 2026-08-06 |
| Weights | Open | Proprietary |
| Context window | — | 524K |
| Max output | — | 131K |
| Input $ / M tokens | — | $0.30 |
| Output $ / M tokens | — | $1.20 |
| Results tracked | 15 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Solar Pro4 leads
DeepSeek LLM 67B: 31.9 (#278), Solar Pro4: 40.1 (#149)
| Benchmark | DeepSeek LLM 67B | Solar Pro4 |
|---|---|---|
| LMArena Coding | 1096 | 1437 |
| LMArena WebDev | — | 1371 |
Reasoning Solar Pro4 leads
DeepSeek LLM 67B: 16.5 (#304), Solar Pro4: 28.5 (#104)
| Benchmark | DeepSeek LLM 67B | Solar Pro4 |
|---|---|---|
| LMArena Hard Prompts | 1070 | 1399 |
| Chess Puzzles | 0% | — |
| Epoch Capabilities Index | 110.5 | — |
Math Solar Pro4 leads
DeepSeek LLM 67B: 8.7 (#324), Solar Pro4: 38.8 (#128)
| Benchmark | DeepSeek LLM 67B | Solar Pro4 |
|---|---|---|
| LMArena Math | 1108 | 1416 |
| OTIS Mock AIME 2024-2025 | 0.8% | — |
| MATH Level 5 | 6.4% | — |
Knowledge Solar Pro4 leads
DeepSeek LLM 67B: 7.0 (#313), Solar Pro4: 39.8 (#129)
| Benchmark | DeepSeek LLM 67B | Solar Pro4 |
|---|---|---|
| GPQA Diamond | 24.6% | — |
| LMArena Expert | — | 1427 |
Multilingual Solar Pro4 leads
DeepSeek LLM 67B: 29.4 (#267), Solar Pro4: 48.7 (#139)
| Benchmark | DeepSeek LLM 67B | Solar Pro4 |
|---|---|---|
| LMArena Non-English | 1073 | 1361 |
| LMArena Chinese | 1132 | 1415 |
| LMArena French | — | 1397 |
| LMArena German | — | 1364 |
| LMArena Japanese | — | 1309 |
| LMArena Korean | — | 1382 |
| LMArena Russian | — | 1360 |
| LMArena Spanish | — | 1401 |
Instruction Following Solar Pro4 leads
DeepSeek LLM 67B: 55.4 (#277), Solar Pro4: 72.7 (#132)
| Benchmark | DeepSeek LLM 67B | Solar Pro4 |
|---|---|---|
| LMArena Instruction Following | 1079 | 1377 |
Long Context Solar Pro4 leads
DeepSeek LLM 67B: 33.1 (#265), Solar Pro4: 42.1 (#130)
| Benchmark | DeepSeek LLM 67B | Solar Pro4 |
|---|---|---|
| LMArena Longer Query | 1092 | 1381 |
Writing & Preference Solar Pro4 leads
DeepSeek LLM 67B: 31.6 (#282), Solar Pro4: 56.5 (#138)
| Benchmark | DeepSeek LLM 67B | Solar Pro4 |
|---|---|---|
| LMArena Text | 1105 | 1386 |
| LMArena Creative Writing | 1067 | 1316 |
| LMArena Multi-Turn | 1082 | 1385 |
Frequently asked questions
Is DeepSeek LLM 67B better than Solar Pro4?
Solar Pro4 is the stronger model overall, scoring 42.1 to 24.9 on the Noometry Index.
Is DeepSeek LLM 67B or Solar Pro4 better for coding?
Solar Pro4 scores higher on coding benchmarks: 40.1 versus 31.9 in the Noometry coding category.
How many benchmarks do DeepSeek LLM 67B and Solar Pro4 share?
10 benchmarks have published results for both models. DeepSeek LLM 67B has 15 scored results on Noometry and Solar Pro4 has 18.