Model comparison
Granite-4.0-350M vs Granite-4.0-H-350M
Head-to-head evidence from 11 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Sibling matchup inside the Granite 4.0 350M family.
Public leaderboard positions: Granite-4.0-350M #181 (Estimated); Granite-4.0-H-350M #182 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. Granite-4.0-350M and Granite-4.0-H-350M share 11 comparable benchmark results. 0 of 8 categories are comparable. 0 results are unique to Granite-4.0-350M; 0 to Granite-4.0-H-350M.
Updated July 23, 2026- Shared results
- 11
- Granite-4.0-350M only
- 0
- Granite-4.0-H-350M only
- 0
- Comparable categories
- 0 / 8
Benchmark data for Granite-4.0-350M and Granite-4.0-H-350M is coming soon on BenchLM.
Confidence note. This is a partial-evidence comparison with 11 shared benchmark results across 5 evidence categories; 0 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
BenchLM has partial data for these models, but not enough overlapping benchmark coverage to produce a fair score-level comparison yet.
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Granite-4.0-350M | Granite-4.0-H-350M | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Granite-4.0-350M$0 input / $0 output | Granite-4.0-H-350M$0 input / $0 output | Listed prices are equal. |
| Generation speedtokens per second | Granite-4.0-350MNot available | Granite-4.0-H-350MNot available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | Granite-4.0-350MNot available | Granite-4.0-H-350MNot available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | Granite-4.0-350M32K | Granite-4.0-H-350M32K | Listed context windows are equal. |
Benchmark Deep Dive
Agentic1 benchmarks
| Benchmark | Granite-4.0-350M | Granite-4.0-H-350M | Result |
|---|---|---|---|
| τ²-bench resultsSource | 13.2% | 14.6% | Granite-4.0-H-350M leads |
Coding1 benchmarks
| Benchmark | Granite-4.0-350M | Granite-4.0-H-350M | Result |
|---|---|---|---|
| AA-SciCodeSource | 0.9% | 1.7% | Granite-4.0-H-350M leads |
Reasoning2 benchmarks
Knowledge6 benchmarks
| Benchmark | Granite-4.0-350M | Granite-4.0-H-350M | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 1.0% | 1.0% | Tie |
| AA-GPQA DiamondSource | 20.3% | 25.7% | Granite-4.0-H-350M leads |
| AA-HLESource | 5.7% | 6.4% | Granite-4.0-H-350M leads |
| AA-Omniscience IndexSource | -72.1% | -87.2% | Granite-4.0-350M leads |
| AA-Omniscience AccuracySource | 3.2% | 3.7% | Granite-4.0-H-350M leads |
| AA-Omniscience Hallucination RateSource | 77.8% | 94.4% | Granite-4.0-350M leads |
Inst. Following1 benchmarks
| Benchmark | Granite-4.0-350M | Granite-4.0-H-350M | Result |
|---|---|---|---|
| AA-IFBenchSource | 16.8% | 17.6% | Granite-4.0-H-350M leads |
Frequently Asked Questions (3)
Can I compare Granite-4.0-350M and Granite-4.0-H-350M on BenchLM yet?
Not fully yet. BenchLM is tracking both models, but the sourced benchmark breakdown for this comparison is still coming soon.
Why does this comparison show “coming soon”?
BenchLM only shows category winners and benchmark-level calls when we have sourced results that can be compared fairly. For these models, the public benchmark coverage is not complete enough yet.
What data is available for Granite-4.0-350M and Granite-4.0-H-350M today?
Granite-4.0-350M: $0.00 input / $0.00 output per 1M tokens Granite-4.0-H-350M: $0.00 input / $0.00 output per 1M tokens Both model pages still include creator, context window, reasoning mode, and other metadata while benchmark coverage fills in.
Related Comparisons
Explore More
Choose a model with this week’s evidence
Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.
One email each week. Unsubscribe anytime.