LLM Comparison

Nemotron 3.5 Lightning 30B A3B vs Trinity Large Thinking

Side-by-side specs, pricing & capabilities

Price vs Intelligence

Add to comparison

2/6 models

Same tier:

	Nemotron 3.5 Lightning 30B A3B	Trinity Large Thinking
Organization	NVIDIA	Arcee AI
OpenTools Score	19 36.2	12 22.6
Family	Nemotron	Trinity
Status	Current	Current
Release Date	Aug 2026	Apr 2026
Context Window	1.0M tokens	262K tokens
Input Price	$0.10/M tokens	$0.22/M tokens
Output Price	$0.95/M tokens	$0.85/M tokens
Pricing Notes	NVIDIA build lists serverless NIM pricing at $0.10 input and $0.95 output per million tokens. Self-hosted weights are also available; infrastructure costs vary.	Cache read: $0.0600/M tokens
Capabilities	textreasoningcodingtool-usefunction-callinglong-contextagenticlocal-deploymentopen-weights	textcodeextended-thinking
Training Cutoff	May 2026 post-training; September 2025 pre-training	—
Max Output	33K tokens	262K tokens
API Identifier	`nvidia/nemotron-3.5-lightning-30b-a3b`	`arcee-ai/trinity-large-thinking`
Benchmarks
Artificial Analysis Intelligence Index v4.3	13.6 indexartificial-analysis	10.9 indexartificial-analysis
Artificial Analysis Intelligence Index v4.3	14 indexhttps://artificialanalysis.ai/models/nemotron-3-5-lightning	—
GPQA Diamond	74.3%artificial-analysis	75.2%artificial-analysis
GPQA Diamond	—	76.3%https://huggingface.co/arcee-ai/Trinity-Large-Thinking
MMLU-Pro	—	83.4%https://huggingface.co/arcee-ai/Trinity-Large-Thinking
SciCode	32.1%artificial-analysis	40.6%artificial-analysis
Terminal-Bench 2.1	24.3%artificial-analysis	20.6%artificial-analysis
	View Nemotron 3.5 Lightning 30B A3B	View Trinity Large Thinking

Cost Calculator

Enter your expected monthly token usage to compare costs.

Input tokens / month

Output tokens / month

Model	Input	Output	Total / mo	vs Best
Nemotron 3.5 Lightning 30B A3BCheapest	$0.10	$0.48	$0.58	—
Trinity Large Thinking	$0.22	$0.43	$0.65	+12%

Compare hosting configurations

Compare two exact setups, including different configurations of the same model. Hardware fit, measured speed and costs are separate decisions.

First configuration

Second configuration

Add your cost assumptions (optional)

Use hours for the same scenario period. Cloud billing includes idle time; active hours alone do not set the bill. Local electricity uses whole-system attributable power, powered hours and your tariff. Missing inputs remain unknown.

Cloud billed hours

Active workload hours

Administration hours

Local powered hours

Attributable power (W)

Electricity rate / kWh (guide currency)

Stored data (GB)

Outgoing data (GB)

Choose two different configurations to see their tradeoffs.

Find configurations for your hardware

NVIDIA

Nemotron 3.5 Lightning 30B A3B

Nemotron 3.5 Lightning is NVIDIA open-weight 30B-A3B reasoning model for fast, long-running agents. Its hybrid Mamba-2, mixture-of-experts, and attention architecture supports function calling, coding, tool use, long context, and efficient local or serverless deployment.

Arcee AI

Trinity Large Thinking

Trinity Large Thinking is a large language model from Arcee AI. Supports up to 262,144 token context window. Available from $0.22/M input tokens.