# IBM Granite 4.2 3B ![Organization](https://img.shields.io/badge/Org-IBM-054ada) ![Params](https://img.shields.io/badge/Params-3B-1f6feb) ![License](https://img.shields.io/badge/License-Apache%202.0-22c55e) ![Released](https://img.shields.io/badge/Released-Aug%202026-555) > **Quick answer:** IBM Granite 4.2 3B is the smallest model in IBM's reasoning-focused Granite 4.2 family — a dense, Apache 2.0 model with a thinking/non-thinking/low-effort switch. It scores 78.3% on AIME25, 67.8% on MMLU-Pro, and supports a native 128K-token context. Unlike its 8B/30B siblings, it does not receive the agentic-RL training stage (no SWE/Terminal/Search environments). ## At a Glance **Where Granite 4.2 3B leads** - Strong reasoning-per-parameter: 78.3% AIME25 and 66.7% HMMT Feb25 at just 3B dense parameters - Native step-by-step reasoning with a **thinking / non-thinking / low-effort** switch — spend more compute only when a task needs it - Fully open under **Apache 2.0** — no licensing restrictions on fine-tuning or production use **Where it lags** - No agentic-RL training — unlike the 8B/30B siblings, it wasn't trained on SWE-agent, terminal, or web-search environments, so it lacks reported scores on SWE-bench, Terminal-Bench, or GDPval - Lower ceiling than its siblings across every reported benchmark (expected at this scale) - 26.7-point gap vs. the 30B model on BFCL (v4) — weaker native tool-calling reliability **Best for:** Lightweight enterprise reasoning/chat workloads (instruction following, math, general QA) where agentic tool-use isn't required — edge or high-throughput deployments where the 8B/30B agentic stack is overkill. ## What Granite 4.2 3B Is Granite 4.2 3B is the smallest of IBM's three-model Granite 4.2 release (3B/8B/30B), the first Granite generation built explicitly around reasoning. All three sizes share the same dense decoder-only transformer architecture and training pipeline — pre-trained from scratch on ~15 trillion tokens across a five-phase strategy, then supervised fine-tuned on chain-of-thought, reasoning, and agentic-trajectory data, then post-trained with a multi-stage GRPO reinforcement-learning pipeline. Where the 3B model differs from its siblings is post-training depth: it completes the **foundational RL** stages (verifiable-reward RL plus instruction-following and code skill boosters) and final RLHF alignment, but takes a shortened path that skips the **agentic RL** block (SWE-agent, Terminal-agent, Search-agent environments) reserved for the 8B and 30B models. Every Granite 4.2 model — including the 3B — supports native tool calling in the OpenAI function-calling format and a `low_effort` thinking mode that spends a short reasoning budget on easy questions. ## Specifications | Field | Value | |---|---| | Organization | IBM | | Parameters | 3B (dense) | | Architecture | Decoder-only dense transformer, GQA (40 attention heads / 8 KV heads, head size 64), RoPE (θ=10,000,000), SwiGLU MLP (hidden size 8192), RMSNorm, 40 layers | | Context window | 131,072 tokens (128K) | | License | Apache 2.0 | | Release date | August 2026 | | Modality | Text only | ## Pricing Open weights available on [Hugging Face](https://huggingface.co/ibm-granite/granite-4.2-3b) (`ibm-granite/granite-4.2-3b`) — free to self-host under Apache 2.0. FP8, NVFP4, MXFP4, and GGUF quantized variants are also published for reduced-memory deployment. ## Context Window Granite 4.2 3B natively supports **131,072 tokens (128K)**. IBM's five-phase pre-training pipeline includes a dedicated long-context phase. ## Public Benchmark Scores | Benchmark | Score | Source | Date | |---|---|---|---| | [BFCL-v4](/benchmarks/bfcl-v4) | 52.41% | Granite 4.2 launch blog (Hugging Face) | 2026-08 | | [AIME 2025](/benchmarks/aime-2025) | 78.33% | Granite 4.2 launch blog (Hugging Face) | 2026-08 | | [HMMT 2025](/benchmarks/hmmt-2025) | 66.67% | Granite 4.2 launch blog (Hugging Face) | 2026-08 | | [LiveCodeBench v6](/benchmarks/livecodebench/livecodebench-v6) | 0.6971 | Granite 4.2 launch blog (Hugging Face) | 2026-08 | | [SciCode](/benchmarks/scicode) | 24.11% | Granite 4.2 launch blog (Hugging Face) | 2026-08 | | [MMLU-Pro](/benchmarks/mmlu-pro) | 67.84% | Granite 4.2 launch blog (Hugging Face) | 2026-08 | | [Arena-Hard-V2](/benchmarks/arena-hard-v2) | 0.3496 | Granite 4.2 launch blog (Hugging Face) | 2026-08 | | [IFBench](/benchmarks/allen-institute-for-ai/ifbench) | 74.33% | Granite 4.2 launch blog (Hugging Face) | 2026-08 | *Scores are self-reported by IBM Research on the official Hugging Face launch blog (huggingface.co/blog/ibm-granite/granite-4-2), which reports results for all three Granite 4.2 sizes in one shared table. IFBench is reported by IBM as "IFBench (prompt)" — the prompt-level-loose submetric, matching Benchgen's `ifbench` metric exactly.* **Benchmarks reported for 8B/30B but not applicable to the 3B model** (no agentic-RL training): SWE-bench Multilingual, SWE-bench Pro, SWE-bench Verified, Terminal-Bench 2.1, Bird-SQL (BirdBench), GDPVal — all reported as "NA" for the 3B size in IBM's own table. **Additional benchmarks reported by IBM, not added to Benchgen:** - **τ³-bench** (45.78) — IBM's table doesn't specify a domain, and Benchgen's only τ³ page (`tau3-banking`) covers a much lower-ceiling banking-specific split (max observed score 33.4 vs. Granite's 45.78) — scale mismatch indicates these aren't the same evaluation; skipped rather than guessed. - **ProfBench** (32.10) — no Benchgen page exists for this benchmark. - **GPQA** (54.80, plain — not "GPQA Diamond") — Benchgen only tracks the Diamond subset; the main GPQA question set is a distinct benchmark and conflating the two risks a wrong-scale/wrong-question-set entry. - **MMLU-ProX lite (IBM)** (27.78) — explicitly labeled an IBM in-house variant, not independently reproducible. - **RULER 64K** (67.52) and **RULER 128K** (55.30) — no Benchgen page exists for RULER yet. ## Related Models & Benchmarks - Compare: [Granite 4.2 8B](/models/ibm/granite-4-2-8b) · [Granite 4.2 30B](/models/ibm/granite-4-2-30b) · [Granite 4.0 Tiny Preview](/models/ibm/granite-4-0-tiny-preview) - Benchmarks: [AIME 2025](/benchmarks/aime-2025) · [MMLU-Pro](/benchmarks/mmlu-pro) · [LiveCodeBench v6](/benchmarks/livecodebench/livecodebench-v6) · [IFBench](/benchmarks/allen-institute-for-ai/ifbench) ## Frequently Asked Questions

What is IBM Granite 4.2 3B?

The smallest model in IBM's Granite 4.2 family — a dense, Apache 2.0 reasoning model with thinking/non-thinking/low-effort modes, built for enterprise chat and instruction-following workloads.

Does Granite 4.2 3B support agentic tool use?

It supports native tool calling, but unlike the 8B and 30B models it wasn't trained through IBM's agentic-RL block (SWE-agent, Terminal-agent, Search-agent), so its agentic task performance is untested/unreported.

Is IBM Granite 4.2 3B open source?

Yes — released under Apache 2.0, IBM's standard Granite license, with no restrictions on fine-tuning or commercial use.

--- *Specs from IBM's official Hugging Face launch blog (huggingface.co/blog/ibm-granite/granite-4-2) and IBM Research blog (research.ibm.com/blog/introducing-granite-4-2). Last updated 2026-08-31.*

IBM Granite 4.2 3B — Training jobs

Training jobs