Updated daily at 05:00 UTC
- Released
- 6 Apr 2026
- Status
- Legacy
- Context
- 131.1K
- Max output
- —
- Input per 1M tokens
- —
- Output per 1M tokens
- —
- Open weights
- Yes
- License
- apache-2.0
- Parameters
- 3.4B
- Knowledge cutoff
- —
- Reasoning
- No
- Modalities
- text → text
Category scores
Out of 100, from independent and third-party results only, by method category-2026-09-29.1.
Not enough independent results yet
Only independent and third-party results rank a model; self-reported and unattributed numbers never do.
Benchmark results
Every result we hold for this model, with who ran it and where it was published.
Knowledge & reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| MMLU-Pro | Accuracy | 49.8% | Self-reported | — | — | Model Card via Hugging Face Community Evals |
Scientific reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| GPQA Diamond | Accuracy | 31.7% | Self-reported | — | — | Model Card via Hugging Face Community Evals |
Other Granite versions
- Granite 4.2 30B7 Aug 2026
- Granite 4.2 3B7 Aug 2026
- Granite 4.2 8B7 Aug 2026
- Granite 4.1 30B6 Apr 2026
- Granite 4.1 8B6 Apr 2026
- Granite 4.0 1B (Nano)7 Oct 2025
- Granite 4.0 350M (Nano)7 Oct 2025
- Granite 4.0 H-1B (Nano)7 Oct 2025
- Granite 4.0 H-350M (Nano)7 Oct 2025
- Granite 4.0 H-Micro (3B)Oct 2025
- Granite 4.0 H-Small (32B-A9B)Oct 2025
- Granite 4.0 H-Tiny (7B-A1B)Oct 2025
- Granite 4.0 Micro (3B)Oct 2025
- Granite 3.3 2B InstructApr 2025
- Granite 3.3 8B InstructApr 2025
- Granite 3.2 2B InstructFeb 2025
- Granite 3.2 8B InstructFeb 2025
- Granite 3.1 1B-A400M InstructDec 2024
- Granite 3.1 2B InstructDec 2024
- Granite 3.1 3B-A800M InstructDec 2024
- Granite 3.1 8B InstructDec 2024
- Granite 3.0 1B-A400M InstructOct 2024
- Granite 3.0 2B InstructOct 2024
- Granite 3.0 3B-A800M InstructOct 2024
- Granite 3.0 8B InstructOct 2024
- Granite 7B Instruct19 May 2024
- Granite 13B Chat / Instruct—