Model comparison
DeepSeek V4 Pro vs Devstral Medium
DeepSeek V4 Pro has enough public results to be ranked (#31); Devstral Medium does not yet, so treat this comparison as directional.
Last verified . 2 shared benchmarks.
DeepSeek
Mistral AISummary
- They share 2 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 2 categories and Devstral Medium in 0 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 21.1.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 53.5% for DeepSeek V4 Pro and 37.7% for Devstral Medium.
- Devstral Medium is cheaper at $0.40 / $2 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 262K.
Side by side
| DeepSeek V4 Pro | Devstral Medium | |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 54.3 | 30.1 |
| Released | 2026-04-24 | 2025-07-10 |
| Weights | Open | Open |
| Context window | 1M | 262K |
| Max output | 393K | 262K |
| Input $ / M tokens | $0.66 | $0.40 |
| Output $ / M tokens | $1.98 | $2 |
| Results tracked | 48 | 2 |
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), Devstral Medium: 24.3
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| LMArena WebDev | 1582 | 1076 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| LMArena Coding | 1470 | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Pro: 32.8 (#58), Devstral Medium: —
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Devstral Medium: 21.1
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| Kagi LLM Benchmark | 53.5% | 37.7% |
| ARC-AGI-2 | 61.3% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| CritPt | 18% | — |
| Chess Puzzles | 47% | — |
| LMArena Hard Prompts | 1461 | — |
| Mystery Game Puzzles | 43% | — |
| DTBench | 93.9% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| Epoch Capabilities Index | 155.31 | — |
| ForecastBench | 56.1 | — |
Math Not comparable
DeepSeek V4 Pro: 64.8 (#30), Devstral Medium: —
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| 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% | — |
| LMArena Math | 1455 | — |
Knowledge Not comparable
DeepSeek V4 Pro: 59.5 (#31), Devstral Medium: —
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| GPQA Diamond | 91.7% | — |
| SimpleQA Verified | 52.9% | — |
| Vectara Hallucination Rate | 8.6% | — |
| LMArena Expert | 1464 | — |
Multilingual Not comparable
DeepSeek V4 Pro: 54.4 (#45), Devstral Medium: —
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| 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), Devstral Medium: —
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| LMArena Instruction Following | 1448 | — |
Long Context Not comparable
DeepSeek V4 Pro: 45.0 (#51), Devstral Medium: —
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| CL-bench Life | 13.5% | — |
| LMArena Longer Query | 1458 | — |
Writing & Preference Not comparable
DeepSeek V4 Pro: 65.5 (#46), Devstral Medium: —
| Benchmark | DeepSeek V4 Pro | Devstral Medium |
|---|---|---|
| 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 Devstral Medium?
DeepSeek V4 Pro has enough public results to be ranked (#31); Devstral Medium does not yet, so treat this comparison as directional.
Which is cheaper, DeepSeek V4 Pro or Devstral Medium?
Devstral Medium is cheaper. It lists at $0.40 per million input tokens and $2 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.
Is DeepSeek V4 Pro or Devstral Medium better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 24.3 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Pro does, with 1M tokens against 262K.
How many benchmarks do DeepSeek V4 Pro and Devstral Medium share?
2 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Devstral Medium has 2.