Key takeaways
- A technical walkthrough of how we built the Granite 4.2 reasoning model family.
- That pipeline includes agentic RL, where the 8B and 30B models learn to act with tools inside real sandboxed environments.
- Every model can produce a chain of thought before its answer and can run in thinking or non-thinking mode depending on how much…
What happened
2 reasoning model family. 2 is our first family of dense, decoder-only reasoning LLMs, released in three sizes: 3B, 8B, and 30B. Each model is pre-trained from scratch on roughly 15T tokens with a five-phase strategy that extends the context window to 512K tokens, supervised fine-tuned on chain-of-thought, reasoning, and agentic-trajectory data, then post-trained with a multi-stage reinforcement learning pipeline.
Phases 1–2 focus on foundational pre-training, phases 3–4 perform mid-training with progressively higher-quality data annealing, and phase 5 introduces long-context training, extending the context window to 512K tokens. Each phase uses a distinct data mixture and learning-rate schedule, gradually shifting from broad web-scale data toward more curated, high-quality sources. 1 blog. Supervised fine-tuning (SFT) turns the base model into a reliable instruction-following, reasoning, and tool-using assistant.
2 million samples, or roughly 100B tokens, of which about 65B are trainable. 2%). These samples and trajectories are generated using a diverse set of agent scaffolds and harnesses, including OpenHands, OpenCode, Terminus-2, SWE-agent, OpenResearcher, MiniSWE, OpenSeeker, EnvScaler, Gemini CLI, Hermes, Codex, and Goose. The agentic data combines samples from both open-source datasets and our own synthetically generated RL environments, spanning a variety of agent–harness combinations. 8%).
We apply multiple stages of quality control before a sample enters the final SFT mixture. First, data from different sources is normalized and reformatted into a consistent OpenAI Chat format, making the conversation structure and tool interactions uniform across datasets and scaffolds. We then use GPT-OSS-120B and Gemma 4 as LLM-based judges to assess sample quality.
Why it matters
That pipeline includes agentic RL, where the 8B and 30B models learn to act with tools inside real sandboxed environments. Every model has a thinking / non-thinking switch, a low-effort thinking mode that spends a short reasoning budget on easy questions, and native tool calling. 0 license. 2 is the reasoning-focused release of the Granite language-model family. 2 adds explicit reasoning.
Every model can produce a chain of thought before its answer and can run in thinking or non-thinking mode depending on how much deliberation a task needs. A low-effort mode falls between the two, spending a short reasoning budget on easy questions.
The three sizes (3B, 8B, and 30B) share the same architectural design and follow the same training pipeline (pre-training from scratch, SFT, then multi-stage RL), each at its own scale. All three are strong reasoners and instruction followers. The clearest capability split shows up in post-training.
The 8B and 30B models additionally go through an agentic RL block that teaches them to operate as agents: calling tools, editing and running code, driving a terminal, and searching the web inside real environments. Every model supports native tool calling. Served through an OpenAI-compatible endpoint (for example, with vLLM), it emits tool calls in the OpenAI function-calling format and plugs into agentic harnesses without extra glue.
2 is also supported in SGLang, see the SGLang cookbook for a ready-to-serve recipe. The rest of this post walks through the build: architecture, pre-training, supervised fine-tuning, the multi-stage RL pipeline, and results. 2 is trained from scratch on approximately 15 trillion tokens using a five-phase training strategy.
What to watch
Low-scoring samples are removed, as are samples containing hallucinated or fabricated information, invalid tool interactions, or tool calls to functions that are not defined in the corresponding tool list. Several targeted, dataset-specific heuristic rules are also applied where appropriate to further improve quality and remove known sources of noise. Finally, we perform both local and global deduplication.
Deduplication is based on SHA-256 hashes computed over the combination of the tools and messages fields, removing duplicate samples both within individual data sources and across the overall SFT mixture. The complete corpus is first globally shuffled to reduce ordering effects and ensure that samples from different domains are well mixed during training.