Arcee AI
4 models built · 1 served
Arcee AI, Inc. is a US company founded in 2023 that builds small language models for enterprise use, best known for the Apache-2.0 licensed Trinity foundation models. Rather than chasing frontier-scale parameter counts, Arcee focuses on models sized for practical deployment and cost control — Trinity Mini and Trinity Large Thinking sit alongside task-specific models like Coder Large and Virtuoso Large. On OpenKey, Arcee AI has 4 models, none free, spanning context windows from 32,768 up to 262,144 tokens. Their newest release, Trinity Large Thinking (April 2026), extends the Trinity line into reasoning-oriented territory while keeping the same open-license approach.
Lineup
For general use with a big context window, Trinity Large Thinking ($0.25/$0.80 per 1M input/output, 262,144 tokens) is the flagship — it's also the only model in the lineup with a listed benchmark result. If you want to cut costs, Trinity Mini runs $0.045/$0.15 per 1M tokens with a 131,072-token window, the cheapest option here by a wide margin. For coding tasks specifically, Coder Large ($0.50/$0.80 per 1M) is purpose-built, while Virtuoso Large ($0.75/$1.20 per 1M) is the priciest model in the lineup, aimed at higher-effort general tasks with the same 131,072-token context as Trinity Mini.
- Trinity Large ThinkingApr 1, 2026 262K $0.258 $0.824 ctx 262K · in $0.258 · out $0.824
- Trinity MiniDec 1, 2025 131K $0.046 $0.154 ctx 131K · in $0.046 · out $0.154
- Coder LargeMay 5, 2025 33K $0.515 $0.824 ctx 33K · in $0.515 · out $0.824
- Virtuoso LargeMay 5, 2025 131K $0.772 $1.24 ctx 131K · in $0.772 · out $1.24
Prices per 1M tokens, flat 3% fee included.
Models Arcee AI serves
Beyond its own models, Arcee AI hosts 1 model in the OpenKey catalog. Prices are Arcee AI’s own list price per 1M input tokens, before our fee.
| Model | Built by | Context | Input /M |
|---|---|---|---|
| Trinity Large Thinking | Arcee AI | 262K | $0.250 |
From the OpenRouter catalog, Jul 28, 2026. All providers
Questions
- What's the price range for Arcee AI models on OpenKey?
- Input pricing runs from $0.045 per 1M tokens (Trinity Mini) up to $0.75 per 1M (Virtuoso Large); output pricing runs from $0.15 to $1.20 per 1M. On OpenKey, add the flat 3% fee: Trinity Mini becomes $0.045 x 1.03 = $0.046 input / $0.15 x 1.03 = $0.155 output, and Virtuoso Large becomes $0.75 x 1.03 = $0.7725 input / $1.20 x 1.03 = $1.236 output.
- Which Arcee AI model should I use if I want the cheapest option?
- Trinity Mini, at $0.045/$0.15 per 1M input/output tokens (provider price), is the cheapest model in Arcee's lineup by a large margin — roughly a fifth the input cost of Coder Large. It still offers a 131,072-token context window, so you're not sacrificing much room for the price.
- Does Arcee AI have any free models on OpenKey?
- No. All 4 Arcee AI models on OpenKey — Trinity Large Thinking, Trinity Mini, Coder Large, and Virtuoso Large — are paid, with none of the platform's 25 free models coming from this lab.