Arcee's Trinity Mini Gets RL-Trained to Run Real Biomedical Research Workflows
Loka post-trained Arcee's open 26B Trinity Mini with RL to run multi-step biomedical research workflows, hitting 81.2% on drug-tool evaluation and 0.863 on biological reasoning
- Loka, Arcee AI, Prime Intellect, and AWS post-trained the open 26B Trinity Mini model with RL for scientific research workflows.
- Two RL environments were used: Drug Tool (evidence-seeking via PubMed, UniProt, GEO, KEGG) and BioReason (structured GO term annotation).
- After 21 controlled training runs, the promoted checkpoint hit 81.2% on Drug Tool and 0.863 on BioReason composite score.
- Training used GRPO + LoRA adapters via Prime Intellect's prime-rl, keeping the base model weights untouched.
- The adapter runs inside a multi-agent scientific app (Strands + FastAPI + React on AWS ECS Fargate) with a critic agent and slash-command workflows.
- All datasets, environments, and code are open-source; Trinity Mini is available via the Arcee API at $0.045/$0.15 per million tokens.
Most AI agents in science are LLMs with a system prompt telling them to use tools. That works until the workflow gets complicated: a protein has three different identifiers, one paper contradicts another, and the model needs to decide what to look up next. A new collaboration between Loka, Arcee AI, Prime Intellect, and AWS actually trained the model to do science rather than prompting it to behave that way.
The result is a post-trained version of Arcee's Trinity Mini, a compact open-weight model fine-tuned through reinforcement learning to execute multi-step biomedical research workflows, coordinate specialized tools, and produce auditable, structured outputs. It now runs inside a full scientific application built by Loka.
A compact model with a demanding job
Trinity Mini is a 26-billion-parameter mixture-of-experts model with only 3 billion active parameters per token, meaning it activates a fraction of its total capacity for any given input, keeping inference fast and cheap. It supports a 128K-token context window across its Nano and Mini variants, enough for long conversations and multi-step workflows without losing coherence, and it was designed for multi-turn agents, tool orchestration, and structured outputs in cloud or on-prem backends.
The team ran 21 controlled post-training experiments using reinforcement learning. The promoted checkpoint, Run 120, is what now powers the deployed agent.
Two training environments, two skills
Training used two distinct RL environments built on Prime Intellect's prime-rl framework, each targeting a different capability:
- Drug Tool RL teaches the model to investigate. Given a biomedical question, it must choose the right tools, recover when retrieval fails, and synthesize a grounded answer. Seven retrieval tools cover PubMed, GEO, KEGG, UniProt, and STRING. When enabled, NVIDIA NIM-backed tools add protein folding, ligand docking, and molecular generation.