IB
IBM
Language model · United States ·Oct 2025

Granite-4.0-H-Small

Language Open (Restricted)
32B
Parameters
21
Benchmarks

About Granite-4.0-H-Small

Granite-4.0-H-Small is an AI model developed by IBM, in the language category, released in 2025, made available as an open-weights model with 32B params.

On this page you'll find Granite-4.0-H-Small's full specifications, including 21 benchmark results. Review provider pricing and benchmark scores below, or compare Granite-4.0-H-Small head-to-head with other language models.

Links

Run Granite-4.0-H-Small locally — what to buy

Share this build ↗

Sized to what Granite-4.0-H-Small actually needs (~8k context at Q4), not to the biggest GPU: Good = cheapest that runs it, Better = best value, Best = most headroom. Speed is a hardware estimate; anything past ~120 tok/s is shown as instant.

Software support

✓ measured · · compatible · — not supported. Informational only — speed is hardware-based.

Community

Would you run Granite-4.0-H-Small again?
No votes yet — be the first.
Did it run? — community reports by hardware

No reports yet. Own this model on some hardware? Be the first to confirm it runs.

Discussion

No comments yet. Start the discussion.

Granite-4.0-H-Small benchmark scores
Benchmark Category Score Variant
HumanEval code 88% pass@1
HumanEval+ code 83% pass@1
MBPP code 84% pass@1
MBPP+ code 71% pass@1
BigCodeBench coding 46.23 pass@1
MultiPL-E coding 57.37 pass@1-multiple
AlpacaEval 2 general 42.48%
ArenaHard general 46.48%
IFEval general 89.87% instruct-strict
IFEval general 87.55% average
IFEval general 85.22% prompt-strict
MMLU general 78.44% 5-shot
MMLU-Pro general 55.47% 5-shot-cot
MMMLU general 69.69% 5-shot
AGIEval knowledge 70.63 0-shot-cot
INCLUDE knowledge 63.97 5-shot
GSM8K math 87.27% 8-shot
MGSM math 38.72 8-shot
BIG-Bench Hard reasoning 81.62% 3-shot-cot
GPQA Diamond reasoning 40.63% 0-shot-cot
BFCL v3 tool-use 64.69%

More models like Granite-4.0-H-Small

Frequently asked questions

What is Granite-4.0-H-Small?

Granite-4.0-H-Small is an AI model developed by IBM, in the language category, released in 2025. It is tracked on GenAIList with its specifications, benchmark scores and provider pricing.

Who created Granite-4.0-H-Small?

Granite-4.0-H-Small was developed by IBM and released in 2025.

Is Granite-4.0-H-Small open source or proprietary?

Granite-4.0-H-Small ships as an open-weights model: you can download and self-host the weights, though the licence may place some restrictions on use.

How much does Granite-4.0-H-Small cost?

Pricing for Granite-4.0-H-Small depends on the provider. See the providers table on this page for the latest API rates.

How does Granite-4.0-H-Small perform on benchmarks?

Granite-4.0-H-Small is benchmarked across 21 evaluations on GenAIList, including HumanEval (88%). See the full benchmark table below and compare it with other models.

IBM이 만든 LLM. Granite-4.0-H-Small. 32B-A9B 성능테스트

Know a great video about Granite-4.0-H-Small?