About
Trinity Large Thinking is a powerful open source reasoning model from the team at Arcee AI. It shows strong performance in PinchBench, agentic workloads, and reasoning tasks. Launch video: https://youtu.be/Gc82AXLa0Rg?si=4RLn6WBz33qT--B7...
Related Models
Pricing
Input
$0.25
per 1M tokens
Output
$0.80
per 1M tokens
Blended
$0.39
per 1M tokens
Cheaper than 60% of models. Median price is $0.67/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$0.39
Monthly
$11.62
vs. Similar Models
DeepSeek: DeepSeek V3.2Q:+0.2
$0.24-38%
MiMo-V2-Flash (Non-reasoning)Q:+0.2
$0.15-61%
Grok 4.3 (Non-reasoning)Q:+0.3
$1.56+303%
Qwen3.6 35B A3B (Non-reasoning)Q:-0.3
$0.84+118%
Performance
211
tokens/sec
Faster than 89% of models
0.71
seconds
Faster than 72% of models
10.17
seconds
Faster than 40% of models
Market Median
93 tok/s
127% faster
Median TTFT
1.13s
37% faster
Throughput/Dollar
546
tok/s per $/1M
Speed Comparison
Nova 2.0 Lite (Non-reasoning)
212 tok/s+0%
OpenAI: GPT-5.1-Codex-Mini
210 tok/s-1%
Google: Gemini 2.5 Flash Lite
215 tok/s+2%
Context Window
262K
tokens
Larger than 60% of models
Max Output
80K
tokens
31% of context
Benchmarks
MMLU-ProNot evaluated
GPQA Diamond
75.2%
HLE
14.7%
LiveCodeBenchNot evaluated
SciCode
36.1%
TerminalBench Hard
22.7%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025Not evaluated
IFBench
56.3%
Long Context Recall
33.0%
Tau2
90.1%
Market AverageTop Score