Model comparison
DeepSeek V4 Pro vs Qwen3 14B
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 35.5 on the Noometry Index. Qwen3 14B costs 1.6× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.
Last verified . 10 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 10 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 6 categories and Qwen3 14B in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 18.5.
- The biggest single-benchmark swing is Chess Puzzles: 47% for DeepSeek V4 Pro and 4% for Qwen3 14B.
- Qwen3 14B is cheaper at $0.35 / $1.40 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 131K.
Side by side
| DeepSeek V4 Pro | Qwen3 14B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 54.3 | 35.5 |
| Released | 2026-04-24 | 2025-04 |
| Weights | Open | Open |
| Context window | 1M | 131K |
| Max output | 393K | 8K |
| Input $ / M tokens | $0.66 | $0.35 |
| Output $ / M tokens | $1.98 | $1.40 |
| Results tracked | 48 | 12 |
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), Qwen3 14B: 37.3 (#195)
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| SciCode | 51% | 31.6% |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| WeirdML | 66.2% | — |
| LMArena Coding | 1470 | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use DeepSeek V4 Pro leads
DeepSeek V4 Pro: 32.8 (#58), Qwen3 14B: 29.6 (#83)
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Berkeley Function Calling Leaderboard | — | 41% |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Qwen3 14B: 18.5 (#280)
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| Kagi LLM Benchmark | 53.5% | 49.1% |
| CritPt | 18% | 0% |
| Chess Puzzles | 47% | 4% |
| DTBench | 93.9% | 64% |
| LMCA | 45.5% | 18.2% |
| Epoch Capabilities Index | 155.31 | 138.23 |
| ARC-AGI-2 | 61.3% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| LMArena Hard Prompts | 1461 | — |
| Mystery Game Puzzles | 43% | — |
| Surface Evolver Bench | 40% | — |
| ForecastBench | 56.1 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Qwen3 14B: 38.6 (#133)
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.6% | 66.4% |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| ProofBench | 50% | — |
| LMArena Math | 1455 | — |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Qwen3 14B: 39.3 (#134)
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| GPQA Diamond | 91.7% | 63.8% |
| Vectara Hallucination Rate | 8.6% | 5.4% |
| SimpleQA Verified | 52.9% | — |
| LMArena Expert | 1464 | — |
Multilingual Not comparable
DeepSeek V4 Pro: 54.4 (#45), Qwen3 14B: —
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| LMArena Non-English | 1439 | — |
| LMArena Chinese | 1486 | — |
| LMArena French | 1472 | — |
| LMArena German | 1458 | — |
| LMArena Japanese | 1445 | — |
| LMArena Korean | 1447 | — |
| LMArena Russian | 1453 | — |
| LMArena Spanish | 1458 | — |
Instruction Following Not comparable
DeepSeek V4 Pro: 76.1 (#47), Qwen3 14B: —
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| LMArena Instruction Following | 1448 | — |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), Qwen3 14B: 38.1 (#204)
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| Fiction.LiveBench | — | 62.5% |
| CL-bench Life | 13.5% | — |
| LMArena Longer Query | 1458 | — |
Writing & Preference Not comparable
DeepSeek V4 Pro: 65.5 (#46), Qwen3 14B: —
| Benchmark | DeepSeek V4 Pro | Qwen3 14B |
|---|---|---|
| LMArena Text | 1451 | — |
| LMArena Creative Writing | 1446 | — |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
| LMArena Multi-Turn | 1467 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Qwen3 14B?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 35.5 on the Noometry Index. Qwen3 14B costs 1.6× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.
Which is cheaper, DeepSeek V4 Pro or Qwen3 14B?
Qwen3 14B is cheaper. It lists at $0.35 per million input tokens and $1.40 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.
Is DeepSeek V4 Pro or Qwen3 14B better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 37.3 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Pro does, with 1M tokens against 131K.
How many benchmarks do DeepSeek V4 Pro and Qwen3 14B share?
10 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Qwen3 14B has 12.