Route difference
The first layer is not brand size. It is user market, protocol entry, model-family coverage, and real call stability.
- NVIDIA: global route, 11 Models
- Liquid: global route, 3 Models
Provider comparison
NVIDIA vs Liquid
Start with the provider route, compare model coverage, then use read-only key checking to verify what is actually callable.
Region
global route / global route
global
Models
11 / 3
56 providers indexed
NVIDIA
Llama 3.1 Nemotron 70B Instruct · Llama 3.1 Nemotron Ultra 253B v1 · Llama 3.3 Nemotron Super 49B V1.5
Sample models
Liquid
LFM2-24B-A2B · LFM2.5-1.2B-Instruct (free) · LFM2.5-1.2B-Thinking (free)
Sample models
Left provider
NVIDIA
nvidia
Regionglobal
Models11
Sample models
Llama 3.1 Nemotron 70B Instruct
Llama 3.1 Nemotron Ultra 253B v1
Llama 3.3 Nemotron Super 49B V1.5
Right provider
Liquid
liquid
Regionglobal
Models3
Sample models
LFM2-24B-A2B
LFM2.5-1.2B-Instruct (free)
LFM2.5-1.2B-Thinking (free)
Comparison summary
How to judge first
NVIDIA and Liquid are both on the global route, so compare protocol, model coverage, and key visibility.
NVIDIA has broader catalog coverage, but broader coverage does not mean every key can call every model.
In production, one key should reveal provider identity, callable models, balance or quota signals, limits, and risk notes.
Key checking value
A key may expose only part of the models, interfaces, or regions, so the final call must rely on detection results, not the catalog alone.
- NVIDIA: Llama 3.1 Nemotron 70B Instruct · Llama 3.1 Nemotron Ultra 253B v1 · Llama 3.3 Nemotron Super 49B V1.5
- Liquid: LFM2-24B-A2B · LFM2.5-1.2B-Instruct (free) · LFM2.5-1.2B-Thinking (free)
Procurement and distribution
For procurement, agency, and marketplace listing, provider comparison must end in supply stability, pricing, monitoring, and model updates.
- Do not treat catalog coverage as real sellability.
- Key checking and continuous monitoring are the final commercial loop.