Home/Models/Llama 3.1 Nemotron Safety Guard 8B v3
MODEL DOSSIER / ENGLISH

Llama 3.1 Nemotron Safety Guard 8B v3

llama-3.1-nemotron-safety-guard-8b-v3

英伟达text2025-10-28

Llama 3.1 Nemotron Safety Guard 8B v3 is a text model from 英伟达. The snapshot records a release date of 2025-10-28. Context window: 128K tokens. Maximum output: 4.096K tokens. Input: text. Output: text. Capabilities recorded: open weights.

Directory snapshot 2026-10-26 · Chinese dossier: 完整中文页

Context128K
Max output4.096K
Inputtext
Outputtext
Official input—
Official output—
Capabilitiesopen weights
Release2025-10-28

Official list price

CNY / 1M TOKENS

The snapshot does not record an official list price for this model. We do not estimate one. Ask the provider or read the relay's own pricing page before budgeting.

Relay coverage

NONE IN SNAPSHOT

None of the sites in our current public snapshot listed this model ID. That does not mean no relay offers it; it means we have no listing to show. If you find one, the detection page can test it.

How to verify this model on a relay

THREE STEPS
  1. Send a short request to the relay and read the model field in the response. It must match the ID you sent; a different ID is a mapping, not necessarily a substitution.
  2. Run the same prompt against a known-good endpoint and compare prompt-token counts. A large, repeatable gap points at a different tokenizer or a different backend.
  3. Run the seven checks. Billing, context retention, error shape, parameter compliance, tools, and streaming each produce evidence you can quote.

Same publisher

英伟达

Questions

FAQ

What is the context window of Llama 3.1 Nemotron Safety Guard 8B v3?

The snapshot records 128,000 tokens.

How much does Llama 3.1 Nemotron Safety Guard 8B v3 cost?

The snapshot records no official price. No relay price is recorded either.

Is a relay really serving Llama 3.1 Nemotron Safety Guard 8B v3?

A model ID in a response is not proof. The seven-check detection compares tokenizer ratios, billing, error shape, parameter compliance, tools, and streaming against the advertised family. Run it on a key you can afford to spend, and treat one run as a spot check, not 24/7 monitoring.

Specifications and prices come from the 2026-10-26 public directory snapshot. We do not publish IQ scores or structured-output claims for models where we have no reliable data.