Model comparison

Trinity Large Thinking vs o3

Capability radar, value bars, and side-by-side posture.

← Back to Compare

Trinity Large Thinking
Not scored
value score
o3
93
value score
Trinity Large Thinking ctx
o3 ctx
200K

Capability radar

Value · context · multimodal · openness · speed posture

Value head-to-head

  • o393

Attribute tape

FieldTrinity Large Thinkingo3
DeveloperArcee AIOpenAI
Context200K
ModalitiesText, Code
OpennessProprietary Model
Speed
Price$10/1M in · $40/1M out
Value score93
SummaryTrinity Large Thinking is Arcee AI's open-weight sparse MoE reasoning model (398B parameters, ~13B active per token) designed for agentic workloads, multi-step tool use, and long-horizon planning. It uses explicit chain-of-thought reasoning traces and agentic RL post-training to produce structured thinking before final outputs.OpenAI reasoning model focused on hard problems.
Open Trinity Large ThinkingOpen o3