Inference
EN
IBM

Granite-4.0-H-1B

Granite-4.0-H-1B is a standard model from IBM in the Granite 4.0 1B family. 11 benchmarks count toward its index, in 6 categories.

  • Open weights
  • Standard
  • Input: text
  • Released 2025-10
RankedEnough results for an index.
IndexAll categories, out of 100.16.3±6.9Median 51.0
CoverageIndex weight that has results.80%
SpeedOutput tokens per second.—Median 67/s
First tokenSeconds to the first output token.—Median 1.3 s
Input / 1MUS dollars per 1M input tokens.—Median $0.60
Output / 1MUS dollars per 1M output tokens.—Median $2.55
ContextMaximum tokens in one request.128KMedian 262K
EloLMArena rating and rank.—Median 1422

Run locally

1.46B parameters
16-bit
4.1 GB
8-bit
2.6 GB
4-bit
1.9 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy fits in 8 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.1 counted15.8—
CodingCode writing and repair.1 counted14.3—
ReasoningLogic problems and puzzles.2 counted24.5218 / 227
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.4 counted11.1267 / 268
MultilingualTasks in many languages.1 counted5.0—
InstructionTasks with strict prompt rules.2 counted17.461 / 63
MathMath problems.None counted——

Results

11 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
Critical Physics TasksReasoning0.0%150 / 214
Artificial Analysis Humanity's Last ExamKnowledge5.0%176 / 222
Instruction-Following EvalInstruction78.5%57 / 70
Evaluating Large Language Models Trained on CodeCoding73.0%18 / 22
Multilingual Grade School MathMultilingual27.5%11 / 13
τ²-Bench Tool-Agent-User EvaluationAgentic19.6%133 / 149
Massive Multitask Language UnderstandingKnowledge59.7%31 / 34
Artificial Analysis Long Context ReasoningReasoning7.0%201 / 217
Massive Multitask Language Understanding ProfessionalKnowledge32.9%70 / 75
Artificial Analysis IFBenchInstruction26.2%138 / 146
Graduate-Level Google-Proof Q&AKnowledge29.7%137 / 141
Artificial Analysis GPQA DiamondKnowledge26.3%197 / 199
BIG-Bench HardReasoning59.7%—

Sources

BenchLMCC BY-NC 4.0

History

5 releases
Index16.8 → 16.3
23 Oct 202612 Nov 2026

Questions

How good is Granite-4.0-H-1B?
Its index is 16.3 out of 100. That is place 280 of 283 ranked models.
What is Granite-4.0-H-1B best at?
Reasoning: 24.5 out of 100, place 218 of 227.
What is the context window of Granite-4.0-H-1B?
128K tokens.
Is Granite-4.0-H-1B open source?
Yes. Its weights are published.
When was Granite-4.0-H-1B released?
October 2025.

Badge

Granite-4.0-H-1B on Inference360: place 280 of 283, index 16.3
HTML
<a href="https://inference360.ai/models/granite-4-0-h-1b"><img src="https://inference360.ai/badges/granite-4-0-h-1b.svg" alt="Granite-4.0-H-1B on Inference360: place 280 of 283, index 16.3" height="20"></a>
Markdown
[![Granite-4.0-H-1B on Inference360: place 280 of 283, index 16.3](https://inference360.ai/badges/granite-4-0-h-1b.svg)](https://inference360.ai/models/granite-4-0-h-1b)

Compare with

MiniCPM5-1B15.5Llama 3.1 8B17.9Granite-4.0-H-350M12.7Gemini 1.5 Flash20.2

More from IBM

All models
Granite-4.0-H-350M12.7Granite-4.0-350M11.9Granite-4.0-1BUnrankedGranite 4.2 30B41.6Granite 4.2 8B37.6