Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Sibling matchup inside the Granite 4.0 350M family.
Granite-4.0-350M
~27
Winner · 3/8 categoriesGranite-4.0-H-350M
~24
0/8 categoriesGranite-4.0-350M· Granite-4.0-H-350M
Granite-4.0-350M makes more sense if instruction following is the priority, while Granite-4.0-H-350M is the cleaner fit if its score, price, or context tradeoffs line up better with your workload.
Granite-4.0-350M and Granite-4.0-H-350M sit in the same Granite 4.0 350M family. This page is less about two unrelated model lineages and more about how the siblings trade off on benchmark shape, token costs, and practical limits like context window.
Granite-4.0-350M has the cleaner overall profile here, landing at 27 versus 24. It is a real lead, but still close enough that category-level strengths matter more than the headline number.
Granite-4.0-350M's sharpest advantage is in instruction following, where it averages 61.6 against 55.4. The single biggest benchmark swing on the page is IFEval, 61.6% to 55.4%.
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 | Granite-4.0-350M | Granite-4.0-H-350M |
|---|---|---|
| Agentic | ||
| Coming soon | ||
| Coding | ||
| HumanEval | 38% | 39% |
| Multimodal & Grounded | ||
| Coming soon | ||
| Reasoning | ||
| BBH | 33.3% | 33.1% |
| KnowledgeGranite-4.0-350M wins | ||
| MMLU | 36.2% | 35.0% |
| GPQA | 26.1% | 24.1% |
| MMLU-Pro | 14.4% | 12.1% |
| Instruction FollowingGranite-4.0-350M wins | ||
| IFEval | 61.6% | 55.4% |
| MultilingualGranite-4.0-350M wins | ||
| MGSM | 16.2% | 14.7% |
| Mathematics | ||
| Coming soon | ||
Granite-4.0-350M and Granite-4.0-H-350M are sibling variants in the Granite 4.0 350M family, so the right pick depends on whether you value the better benchmark line, cheaper tokens, or the larger context window. Granite-4.0-350M is ahead overall 27 to 24.
Granite-4.0-350M has the edge for knowledge tasks in this comparison, averaging 18.5 versus 16.4. Inside this category, MMLU-Pro is the benchmark that creates the most daylight between them.
Granite-4.0-350M has the edge for instruction following in this comparison, averaging 61.6 versus 55.4. Inside this category, IFEval is the benchmark that creates the most daylight between them.
Granite-4.0-350M has the edge for multilingual tasks in this comparison, averaging 16.2 versus 14.7. Inside this category, MGSM 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.