Model comparison
DeepSeek V4 Pro vs Olmo 3.1 32b Think
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 37.9 on the Noometry Index.
Last verified . 15 shared benchmarks.
DeepSeek
Allen Institute for AI (Ai2)Summary
- They share 15 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 8 categories and Olmo 3.1 32b Think in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 25.2.
Side by side
| DeepSeek V4 Pro | Olmo 3.1 32b Think | |
|---|---|---|
| Provider | DeepSeek | Allen Institute for AI (Ai2) |
| Noometry Index | 54.3 | 37.9 |
| Released | 2026-04-24 | — |
| Weights | Open | Open |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.66 | — |
| Output $ / M tokens | $1.98 | — |
| Results tracked | 48 | 15 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DeepSeek V4 Pro: 52.4 (#34), Olmo 3.1 32b Think: 37.7 (#189)
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| LMArena Coding | 1470 | 1291 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Pro: 32.8 (#58), Olmo 3.1 32b Think: —
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Olmo 3.1 32b Think: 25.2 (#150)
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| LMArena Hard Prompts | 1461 | 1272 |
| ARC-AGI-2 | 61.3% | — |
| Kagi LLM Benchmark | 53.5% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| CritPt | 18% | — |
| Chess Puzzles | 47% | — |
| Mystery Game Puzzles | 43% | — |
| DTBench | 93.9% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| Epoch Capabilities Index | 155.31 | — |
| ForecastBench | 56.1 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Olmo 3.1 32b Think: 36.3 (#168)
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| LMArena Math | 1455 | 1305 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| OTIS Mock AIME 2024-2025 | 98.6% | — |
| ProofBench | 50% | — |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Olmo 3.1 32b Think: 35.7 (#181)
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| LMArena Expert | 1464 | 1295 |
| GPQA Diamond | 91.7% | — |
| SimpleQA Verified | 52.9% | — |
| Vectara Hallucination Rate | 8.6% | — |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), Olmo 3.1 32b Think: 38.1 (#231)
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| LMArena Non-English | 1439 | 1209 |
| LMArena Chinese | 1486 | 1242 |
| LMArena French | 1472 | 1260 |
| LMArena German | 1458 | 1262 |
| LMArena Russian | 1453 | 1193 |
| LMArena Spanish | 1458 | 1289 |
| LMArena Japanese | 1445 | — |
| LMArena Korean | 1447 | — |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.1 (#47), Olmo 3.1 32b Think: 65.6 (#218)
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| LMArena Instruction Following | 1448 | 1247 |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), Olmo 3.1 32b Think: 38.6 (#195)
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| LMArena Longer Query | 1458 | 1272 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), Olmo 3.1 32b Think: 46.2 (#220)
| Benchmark | DeepSeek V4 Pro | Olmo 3.1 32b Think |
|---|---|---|
| LMArena Text | 1451 | 1272 |
| LMArena Creative Writing | 1446 | 1226 |
| LMArena Multi-Turn | 1467 | 1252 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Olmo 3.1 32b Think?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 37.9 on the Noometry Index.
Is DeepSeek V4 Pro or Olmo 3.1 32b Think better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 37.7 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Pro and Olmo 3.1 32b Think share?
15 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Olmo 3.1 32b Think has 15.