open-source-llm reasoning tool-use agentic-ai local-inference

Granite 4.2 adds reasoning and tool use to dense models

Three open-source reasoning models (3B/8B/30B) trained with 15T tokens, chain-of-thought, and agentic RL in sandboxed environments—8B and 30B learn tool calling and code execution.

Summary

Native tool calling in OpenAI-compatible format and agentic RL training mean you can run self-contained reasoning agents locally without custom glue code. 512K context window supports long-horizon tasks.

Why it matters

Implementation verdict

Replaces GPT-4o for reasoning workflows where you control the compute budget. Requires vLLM or SGLang for serving; agentic features (8B/30B only) need sandboxed tool environments. Try 8B now if you have inference infrastructure; 3B is viable for latency-critical edge cases.

Sources

1.Granite 4.2 is our first family of dense, decoder-only reasoning LLMs, released in three sizes: 3B, 8B, and 30B
2.pre-trained from scratch on roughly 15T tokens
3.extends the context window to 512K tokens
4.the 8B and 30B models learn to act with tools inside real sandboxed environments
5.All Granite 4.2 models are released under the Apache 2.0 license
6.emits tool calls in the OpenAI function-calling format

Dev Signal

Get briefs like this in your inbox — free, every weekday.

100+ sources compressed into one 4-minute read. Ranked, cited, implementation-ready.

Read the full issue →All briefs