Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
DeepSeek V3
51
Winner · 2/8 categoriesGranite-4.0-H-350M
~24
0/8 categoriesDeepSeek V3· Granite-4.0-H-350M
Pick DeepSeek V3 if you want the stronger benchmark profile. Granite-4.0-H-350M only becomes the better choice if you want the cheaper token bill.
DeepSeek V3 is clearly ahead on the aggregate, 51 to 24. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
DeepSeek V3's sharpest advantage is in knowledge, where it averages 57.5 against 16.4. The single biggest benchmark swing on the page is MMLU-Pro, 75.9% to 12.1%.
DeepSeek V3 is also the more expensive model on tokens at $0.27 input / $1.10 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Granite-4.0-H-350M. That is roughly Infinityx on output cost alone. DeepSeek V3 gives you the larger context window at 128K, compared with 32K for Granite-4.0-H-350M.
BenchLM keeps the benchmark table and the operator tradeoffs on the same page so a better score does not hide a materially slower, pricier, or smaller-context model.
Runtime metrics show N/A when BenchLM does not have a sourced snapshot for that exact model. The scoring rules and freshness policy are documented on the methodology page.
| Benchmark | DeepSeek V3 | Granite-4.0-H-350M |
|---|---|---|
| Agentic | ||
| Coming soon | ||
| Coding | ||
| LiveCodeBench | 37.6% | — |
| SWE-bench Verified | 42% | — |
| HumanEval | — | 39% |
| Multimodal & Grounded | ||
| Coming soon | ||
| Reasoning | ||
| LongBench v2 | 48.7% | — |
| BBH | — | 33.1% |
| KnowledgeDeepSeek V3 wins | ||
| GPQA | 59.1% | 24.1% |
| MMLU-Pro | 75.9% | 12.1% |
| SimpleQA | 24.9% | — |
| MMLU | — | 35.0% |
| Instruction FollowingDeepSeek V3 wins | ||
| IFEval | 86.1% | 55.4% |
| Multilingual | ||
| MGSM | — | 14.7% |
| Mathematics | ||
| AIME 2024 | 39.2% | — |
| MATH-500 | 90.2% | — |
DeepSeek V3 is ahead overall, 51 to 24. The biggest single separator in this matchup is MMLU-Pro, where the scores are 75.9% and 12.1%.
DeepSeek V3 has the edge for knowledge tasks in this comparison, averaging 57.5 versus 16.4. Inside this category, MMLU-Pro is the benchmark that creates the most daylight between them.
DeepSeek V3 has the edge for instruction following in this comparison, averaging 86.1 versus 55.4. Inside this category, IFEval is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.