Trinity-Large-Preview

Open Source

400B

Parameters

Benchmarks

About Trinity-Large-Preview

Trinity-Large-Preview is an AI model developed by Arcee AI, released in 2026, made available as an open-source model with 400B params. Trinity-Large-Preview is the chat-ready, lightly post-trained 400B sparse Mixture-of-Experts checkpoint from Arcee AI — .

On this page you'll find Trinity-Large-Preview's full specifications, including 2 benchmark results. Review provider pricing and benchmark scores below, or compare Trinity-Large-Preview head-to-head with other AI models.

→ More models by Arcee AI → AI benchmarks leaderboard

Links

Announcement ↗ Paper ↗ Hugging Face ↗

Community

Would you run Trinity-Large-Preview again?

No votes yet — be the first.

Did it run? — community reports by hardware

No reports yet. Own this model on some hardware? Be the first to confirm it runs.

Discussion

No comments yet. Start the discussion.

Trinity-Large-Preview benchmark scores▶

Compare on the benchmark table →

Benchmark	Category	Score	Variant
HumanEval	code	84.8%	pass@1
MMLU	general	82.3%

More models like Trinity-Large-Preview

Trinity-Large-Thinking

Arcee AI · 400B

Frequently asked questions

What is Trinity-Large-Preview? ▶

Trinity-Large-Preview is the chat-ready, lightly post-trained 400B sparse Mixture-of-Experts checkpoint from Arcee AI — It is tracked on GenAIList with its specifications, benchmark scores and provider pricing.

Who created Trinity-Large-Preview? ▶

Trinity-Large-Preview was developed by Arcee AI and released in 2026.

Is Trinity-Large-Preview open source or proprietary? ▶

Trinity-Large-Preview is released as open source — its weights are publicly available to download, run locally and fine-tune.

How much does Trinity-Large-Preview cost? ▶

Pricing for Trinity-Large-Preview depends on the provider. See the providers table on this page for the latest API rates.

How does Trinity-Large-Preview perform on benchmarks? ▶

Trinity-Large-Preview is benchmarked across 2 evaluations on GenAIList, including HumanEval (84.8%). See the full benchmark table below and compare it with other models.