Model comparison
DeepSeek V4 Pro vs Mixtral 8x22B
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 27.1 on the Noometry Index.
Last verified . 22 shared benchmarks.
DeepSeek
Mistral AISummary
- They share 22 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 9 categories and Mixtral 8x22B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek V4 Pro leads 59.5 to 15.1.
- The biggest single-benchmark swing is WeirdML: 66.2% for DeepSeek V4 Pro and 3.2% for Mixtral 8x22B.
- DeepSeek V4 Pro is cheaper at $0.66 / $1.98 per million input/output tokens, against $2 / $6 for Mixtral 8x22B.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 64K.
Side by side
| DeepSeek V4 Pro | Mixtral 8x22B | |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 54.3 | 27.1 |
| Released | 2026-04-24 | 2024-04-17 |
| Weights | Open | Open |
| Context window | 1M | 64K |
| Max output | 393K | 64K |
| Input $ / M tokens | $0.66 | $2 |
| Output $ / M tokens | $1.98 | $6 |
| Results tracked | 48 | 34 |
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), Mixtral 8x22B: 24.2 (#329)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| WeirdML | 66.2% | 3.2% |
| LMArena Coding | 1470 | 1166 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 51% | — |
| BigCodeBench Instruct | — | 40.6% |
| BigCodeBench Complete | — | 50.2% |
| ALE-Bench | 1,403 | — |
| HumanEval+ | — | 72% |
| MBPP+ | — | 64.3% |
Agentic & Tool Use DeepSeek V4 Pro leads
DeepSeek V4 Pro: 32.8 (#58), Mixtral 8x22B: 23.1 (#127)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Cybench | — | 7.5% |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Mixtral 8x22B: 19.9 (#248)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| LMArena Hard Prompts | 1461 | 1150 |
| DTBench | 93.9% | 55.1% |
| Epoch Capabilities Index | 155.31 | 122.03 |
| ForecastBench | 56.1 | 56.3 |
| 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% | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Mixtral 8x22B: 22.9 (#275)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| LMArena Math | 1455 | 1184 |
| 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% | — |
| Omni-MATH | — | 16.3% |
| MATH Level 5 | — | 24.2% |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Mixtral 8x22B: 15.1 (#293)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| GPQA Diamond | 91.7% | 34.1% |
| LMArena Expert | 1464 | 1113 |
| SimpleQA Verified | 52.9% | — |
| MMLU-Pro | — | 46% |
| Vectara Hallucination Rate | 8.6% | — |
| GPQA (HELM) | — | 33.4% |
| MMLU | — | 77.8% |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), Mixtral 8x22B: 32.8 (#255)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| LMArena Non-English | 1439 | 1128 |
| LMArena Chinese | 1486 | 1116 |
| LMArena French | 1472 | 1166 |
| LMArena German | 1458 | 1141 |
| LMArena Japanese | 1445 | 1037 |
| LMArena Korean | 1447 | 1057 |
| LMArena Russian | 1453 | 1158 |
| LMArena Spanish | 1458 | 1151 |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.1 (#47), Mixtral 8x22B: 57.7 (#266)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| LMArena Instruction Following | 1448 | 1147 |
| IFEval | — | 72.4% |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), Mixtral 8x22B: 34.7 (#247)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| LMArena Longer Query | 1458 | 1144 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), Mixtral 8x22B: 36.9 (#262)
| Benchmark | DeepSeek V4 Pro | Mixtral 8x22B |
|---|---|---|
| LMArena Text | 1451 | 1162 |
| LMArena Creative Writing | 1446 | 1141 |
| LMArena Multi-Turn | 1467 | 1130 |
| EQ-Bench Creative Writing | 1553 | — |
| WildBench | — | 71.1% |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Mixtral 8x22B?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 27.1 on the Noometry Index.
Which is cheaper, DeepSeek V4 Pro or Mixtral 8x22B?
DeepSeek V4 Pro is cheaper. It lists at $0.66 per million input tokens and $1.98 per million output tokens; Mixtral 8x22B lists at $2 and $6.
Is DeepSeek V4 Pro or Mixtral 8x22B better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 24.2 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Pro does, with 1M tokens against 64K.
How many benchmarks do DeepSeek V4 Pro and Mixtral 8x22B share?
22 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Mixtral 8x22B has 34.