Model comparison
DeepSeek V4 Pro vs Ministral 8B
DeepSeek V4 Pro is the stronger model overall, scoring 54.2 to 27.8 on the Noometry Index. Ministral 8B costs 5.4× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.
Last verified . 15 shared benchmarks.
DeepSeek
Mistral AISummary
- They share 15 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 9 categories and Ministral 8B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek V4 Pro leads 59.5 to 12.5.
- The biggest single-benchmark swing is GPQA Diamond: 91.7% for DeepSeek V4 Pro and 27.1% for Ministral 8B.
- Ministral 8B is cheaper at $0.10 / $0.10 per million input/output tokens, against $0.43 / $0.87 for DeepSeek V4 Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 128K.
Side by side
| DeepSeek V4 Pro | Ministral 8B | |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 54.2 | 27.8 |
| Released | 2026-04-24 | 2024-10-01 |
| Weights | Open | Open |
| Context window | 1M | 128K |
| Max output | 393K | 128K |
| Input $ / M tokens | $0.43 | $0.10 |
| Output $ / M tokens | $0.87 | $0.10 |
| Results tracked | 48 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DeepSeek V4 Pro: 52.2 (#31), Ministral 8B: 34.4 (#226)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| LMArena Coding | 1470 | 1202 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use DeepSeek V4 Pro leads
DeepSeek V4 Pro: 32.7 (#54), Ministral 8B: 16.2 (#140)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Berkeley Function Calling Leaderboard | — | 11.1% |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.4 (#22), Ministral 8B: 17.7 (#279)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| LMArena Hard Prompts | 1461 | 1191 |
| DTBench | 93.9% | 45.7% |
| 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% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| Epoch Capabilities Index | 155.39 | — |
| ForecastBench | 56.1 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.6 (#24), Ministral 8B: 23.7 (#262)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| LMArena Math | 1455 | 1188 |
| 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% | — |
| MATH Level 5 | — | 14.9% |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#28), Ministral 8B: 12.5 (#292)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| GPQA Diamond | 91.7% | 27.1% |
| Vectara Hallucination Rate | 8.6% | 7.4% |
| LMArena Expert | 1463 | 1170 |
| SimpleQA Verified | 52.9% | — |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.5 (#44), Ministral 8B: 35.3 (#242)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| LMArena Non-English | 1438 | 1165 |
| LMArena Chinese | 1486 | 1193 |
| LMArena Russian | 1452 | 1195 |
| LMArena French | 1473 | — |
| LMArena German | 1458 | — |
| LMArena Japanese | 1448 | — |
| LMArena Korean | 1452 | — |
| LMArena Spanish | 1457 | — |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.3 (#44), Ministral 8B: 60.7 (#245)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| LMArena Instruction Following | 1448 | 1161 |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#49), Ministral 8B: 36.7 (#222)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| LMArena Longer Query | 1459 | 1212 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.6 (#45), Ministral 8B: 39.8 (#241)
| Benchmark | DeepSeek V4 Pro | Ministral 8B |
|---|---|---|
| LMArena Text | 1451 | 1191 |
| LMArena Creative Writing | 1446 | 1175 |
| LMArena Multi-Turn | 1467 | 1166 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Ministral 8B?
DeepSeek V4 Pro is the stronger model overall, scoring 54.2 to 27.8 on the Noometry Index. Ministral 8B costs 5.4× 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 Ministral 8B?
Ministral 8B is cheaper. It lists at $0.10 per million input tokens and $0.10 per million output tokens; DeepSeek V4 Pro lists at $0.43 and $0.87.
Is DeepSeek V4 Pro or Ministral 8B better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.2 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Pro does, with 1M tokens against 128K.
How many benchmarks do DeepSeek V4 Pro and Ministral 8B share?
15 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Ministral 8B has 17.