Arcee AI

Trinity Large Thinking

Trinity Large Thinking — Arcee reasoning-optimized sparse MoE model with extended chain-of-thought and 256K context.

trinity-large-thinking

Catalog preview. Paid requests for this model are not enabled yet.

List pricing

—

USD. No uRouter markup. Live prices can change.

Input: $0.3 / 1,000,000 tokens
Output: $1 / 1,000,000 tokens
Cache read: $0.045 / 1,000,000 tokens
Cache write: $0.3 / 1,000,000 tokens

How pricing works

Model details

Input: text
Output: text
Context window: 256,000 tokens
Maximum output: 32,768 tokens
Availability: Unknown
Updated: 11/8/2026, 1:45:40 AM UTC

Capacity is reserved when your request is accepted.

Capabilities

chatstreamingtoolsreasoning

Supported parameters

max_tokenstemperaturetop_pfrequency_penaltypresence_penaltyrepetition_penaltymin_plogit_biaslogprobstop_logprobsstoptoolstool_choiceresponse_formatseedstructured_outputsinclude_reasoningreasoning

Listed capabilities describe the model. uRouter execution support is verified separately.