whatmodel.sh / Trinity-Large-Thinking

Trinity-Large-Thinking

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Trinity-Large-Thinking is an open-weights language model from Arcee AI, released April 2026. Its published size is 398B parameters (13B active per token) and a 512k-token context window. Licence: openmdw-1.1.

Catalogue identity notes

Arena entry 'trinity-large-thinking' (text, webdev). Maker repo arcee-ai/Trinity-Large-Thinking @dc6a99b: licence tag other (openmdw-1.1). Parameters: card: 'Total parameters ~398B', 'Active parameters per token ~13B'. Released 2026-04: Arcee blog 'Trinity-Large-Thinking' dated Apr 1, 2026. Reasoning-optimised Trinity Large checkpoint; listed on Arcee's API pricing page.

62.6±1.6#170 of 719 LLMsfrom 5 benchmark results across 2 sources

$ how it compares

The LLMs ranked either side of Trinity-Large-Thinking, out of 828 with a score, then more from Arcee AI.

#modelscorewhereRAMreleased
197Seed-OSS 36B Instruct ByteDance Seed63±1local~23GB2025-08
198GPT-4.5 Preview OpenAI63±1comparison only2025-02
199Trinity-Large-Thinking Arcee AI63±2api $0.39/M +local~141GB2026-04
200Magistral Medium 1.2 (2509) Mistral63±2comparison only2025-09
201MiniMax-M1 (80k) MiniMax62±1api $0.96/M +local~959GB2025-06
more from Arcee AI
249Trinity-Large-Preview Arcee AI59±2local~155GB2026-01

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasksnot ranked on this task yet · 1 test
testresultvs the others in its rundifficultycountswhy
τ²-bench Telecom lab94.7% ✎▁·▅█▆▅▂▃▃▂▃▁ beats 74/89very hard · avg 9.6%—not counted yet: no other model in the same run to compare it with
coding▁▁▁▃▆██▇▆▄▃▁ #169 of 458 * · 43% of its score · 1 test

Beats Qwen3-Coder-Next (80B-A3B), Qwen3-VL 235B-A22B Instruct, Qwen3.5 9B here, though they rank higher overall. Trails QwQ-32B-Preview, Ministral 3 8B Reasoning 2512, Phi-4-reasoning 14B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena WebDev (Code Arena)1237.6 ✎▁▂▄▃▅█▄▇▆▄▁▁ beats 9/119avg 981▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 43%sharp
chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #141 of 393 * · 24% of its score · 1 test

Beats INTELLECT-3, Nemotron 3 Super 120B-A12B, MiniMax-M2 here, though they rank higher overall. Trails LFM2.5 1.2B Thinking, LFM2.5 8B-A1B, LFM2.5 2.6B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
LMArena (text)1367.3 ✎▁▁▁▂▃▃▃▆▆▇█▃ beats 211/379avg 1268▇▇▇▇▇▇▇▇▇▇ 24%blunt
math█··········█ #1 of 2 * · 3.2% of its score · 1 test
testresultvs the others in its rundifficultycountswhy
AIME 2025 lab96.3% ✎▁·▁▁▁▂▅▆█▇▅▃ beats 132/133medium · avg 39%▇ 3.2%sharp · noisy · small run
other tests2 not in a task · 28% of its score
testresultvs the others in its rundifficultycountswhy
GPQA Diamond lab76.3% ✎▁▁▁▁▂▄▅▅███▄ beats 180/310hard · avg 44%▇▇▇▇▇▇▇▇ 19%sharp · at its level · small run
MMLU-Pro lab83.4% ✎▁▁·▁▁▁▃▂▄▅█▁ beats 156/188medium · avg 59%▇▇▇▇ 9.2%small run

$ where to run it

Through an API

api · $0.39/M · $0.25 in / $0.8 out per 1M tokens · 512k context · docs.arcee.ai ↗ · price sourceUSD per 1M tokens, Arcee Platform text-model pricing table, model trinity-large-thinking: input $0.25, output $0.80; cached input $0.06 excluded. Currency: USD. Arcee official pricing docs, observed 2026-10-07.

llama.cpp · recommended

version

  1. Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.
    macOS / Linux
    curl -LsSf https://llama.app/install.sh | sh
    Windows
    irm https://llama.app/install.ps1 | iex

    other ways to install ↗

  2. Run Trinity-Large-Thinking · needs ~241.85 GB · from huggingface.co/arcee-ai (official)
    chat in the terminal
    llama cli -hf arcee-ai/Trinity-Large-Thinking-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf arcee-ai/Trinity-Large-Thinking-GGUF:Q4_K_M

LM Studio

version

  1. Install LM Studio · Open the app once before using the lms command.

    download it (macOS / Windows / Linux) ↗

  2. Run Trinity-Large-Thinking · needs ~241.85 GB · from huggingface.co/arcee-ai (official)

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/arcee-ai/Trinity-Large-Thinking-GGUF@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Official weights and documentation: huggingface.co/arcee-ai ↗

$ how the score was worked out · LMArena leaderboard dataset 67%, Arcee AI vendor-reported benc… 32%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
LMArena leaderboard dataset2▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 67%CC BY 4.0
Arcee AI vendor-reported benchmark scores for its own Trinity models (Hugging Face model cards) · lab-reported3▇▇▇▇▇▇▇▇▇ 32%Factual vendor-reported score (not licensed content)
prior · pull toward the middle▇ 1%
weighted average= 62.6

not counted (1)

Arcee AI vendor-reported benchmark scores for its own Trinity models (Hugging Face model cards)τ²-bench Telecom [correct]94.7%fewer than three models have this test yet, or this source ran it for this model alone

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json