Skip to content
Apertis
Get started
NVIDIAChat

Llama 3.1 Nemotron Ultra 253B v1 (Free)

llama-3.1-nemotron-ultra-253b-v1:free

Llama 3.1 Nemotron Ultra 253B v1 (Free) by NVIDIA.

Context
131.1K tokens
Endpoint
Get API KeyCompare

Pricing

Input$0 / 1M
Output$0 / 1M
Cache Write$0 / 1M
Cache Read$0 / 1M
Prompt cache writes and reads are included at no additional cost.

Quick Start

Select an endpoint and copy a working example for this model.

Endpoint:
python
from openai import OpenAI client = OpenAI(    api_key="YOUR_API_KEY",    base_url="https://api.apertis.ai/v1") response = client.chat.completions.create(    model="llama-3.1-nemotron-ultra-253b-v1:free",    messages=[        {"role": "user", "content": "Hello!"}    ],    max_tokens=1024,    temperature=0.7) print(response.choices[0].message.content) # Optional: Enable context compression to reduce token usage# response = client.chat.completions.create(#     model="llama-3.1-nemotron-ultra-253b-v1:free",#     messages=[{"role": "user", "content": "Hello!"}],#     extra_body={"compression": {"enabled": True, "model": "gpt-4.1-mini"}}# )

Supported Parameters

API docs
Common7 params
modelmessagesmax_tokenstemperaturetop_pstreamtools
Extended4 params
reasoning_effortstream_optionsthinkingextra_body

Cursor IDE Model IDs

Use these namespaced identifiers in Cursor IDE to avoid conflicts with built-in models.

llama-3.1-nemotron-ultra-253b-v1:free

Decision guidance

Fit the request contract before production.

Multi-provider fallback guide

Use this model when

  • The workload fits the text or chat tasks shown in this model's current catalog record.
  • The input fits within the published 131.1K-token context record, with output and system overhead budgeted separately.

Check before production

  • Confirm the production client uses one of the listed request surfaces: /v1/chat/completions, /v1/responses, /v1/messages.
  • Estimate a representative request from the current Free price fields instead of extrapolating from a tiny prompt.
  • Check the observed-availability card and run your own timeout, retry, and fallback test before relying on the model.

Compare with Other Models

See how this model compares to others from the same provider.

Nemotron 3.5 Lightning

NVIDIA Nemotron 3.5 Lightning is an open Mixture-of-Experts (MoE) model with 30B total parameters and 3B active per token, optimized for high-throughput agentic workloads and efficient inference. Its lightweight active compute and open design make it well suited for specialized agents, domain-specific customization, and scalable production deployments where speed, cost efficiency, and adaptability are key.

Context
1M
Input
$0/M
Output
$0/M

Nemotron 3.5 Lightning (Free)

NVIDIA Nemotron 3.5 Lightning is an open Mixture-of-Experts (MoE) model with 30B total parameters and 3B active per token, optimized for high-throughput agentic workloads and efficient inference. Its lightweight active compute and open design make it well suited for specialized agents, domain-specific customization, and scalable production deployments where speed, cost efficiency, and adaptability are key.

Context
1M
Input
$0/M
Output
$0/M

Nemotron 3 Nano Omni (Free)

NVIDIA Nemotron 3 Nano Omni is an open 30B-A3B multimodal model designed as a perception and context sub-agent for enterprise agent systems. It supports text, image, video, and audio inputs with text output, enabling unified multimodal reasoning within a single inference loop. Built on a hybrid MoE Transformer–Mamba architecture with Conv3D video layers and Efficient Video Sampling (EVS), it delivers significantly improved efficiency for video reasoning—achieving ~2× higher throughput and 2.5× lower compute compared to separate pipelines. With up to 300K context length and extended thinking support, it is well suited for scalable, multimodal agent workflows.

Context
256K
Input
$0/M
Output
$0/M

Llama 3.3 Nemotron Super 49B V1.5

Llama-3.3-Nemotron-Super-49B-v1.5 is a 49B reasoning and chat model derived from Llama-3.3-70B-Instruct, tuned for agent workflows like RAG and tool calling with a 128K context window. It combines supervised training with multiple RL stages to improve alignment, step-by-step reasoning, and tool use, while a NAS “Puzzle” architecture reduces memory and boosts throughput so it can run on a single H100/H200. It delivers strong results across math and coding benchmarks, supports toggleable reasoning modes, and is designed for efficient, reliable agent systems and long-context retrieval where accuracy and cost balance matter.

Context
131.1K
Input
$0.05/M
Output
$0.20/M