Tasks and inputs
Trinity Mini is listed for Text generation. It accepts text and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
Trinity Mini is Arcee AI's open-weight mixture-of-experts language model. It has 26 billion total parameters, with about 3 billion activated per token.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
Trinity Mini is Arcee AI's open-weight mixture-of-experts language model. It has 26 billion total parameters, with about 3 billion activated per token.
The active count describes a part of the token-level computation. The total count still matters when storing and loading the weights. Do not budget hardware as though this were a dense three-billion-parameter checkpoint. Measure memory use with your chosen precision, context length and batch size.
The published configuration allows 131,072 context positions. The model card documents several serving options, including vLLM and a hosted route. Tool calling and reasoning output need the appropriate parser in the selected runtime. A model loading successfully is only the first compatibility check.
Test code against executable checks and tool calls against a strict schema. Count incorrect arguments, unsupported assumptions and repair attempts. For retrieval or document work, include questions whose answers are absent from the input; a useful model should not fill every gap with an invention.
The checkpoint uses the OpenMDW-1.1 license. Review that license for your deployment rather than assuming every open-weight model has identical terms. Browse Arcee AI models, or use the comparison tool to inspect another candidate.
Compare the model on the work you need to complete.
Trinity Mini is listed for Text generation. It accepts text and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
Documented limits: 131.07K context. Measure a completed task using the prompts, documents and tools your application will use. Include reasoning and retries in the result.
The weights are downloadable under OpenMDW-1.1. Check runtime support and size the complete model and working context for your expected concurrency.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
Anthropic
Claude’s small model for high-volume extraction, classification and routing, with adaptive thinking and a 1M-token context window.
Mistral AI
Mistral Large 4 is an open-weight multimodal model released in public preview on the publisher’s announced release, for general reasoning, coding and agent applications.
OpenAI
GPT-6.1 Sol is OpenAI's updated Sol model for coding and professional tasks. It accepts text and images and returns text.
inclusionAI
Ling 3.1 Flash is inclusionAI’s hybrid reasoning model for coding and tool agents, available through a hosted AI Gateway route.
Anthropic
Claude Sonnet 5.5 is an Anthropic language model that accepts text and images and returns text.
MiniMax
MiniMax M3.1 Flash Preview is a multimodal reasoning model for code and agent workflows, accessed through eligible MiniMax plans and tools.