impulse

Nemotron 3 Ultra 550B

Open weights
API id nemotron-3-ultra-550b
Input / 1M
$0.675
at Kyma API
Output / 1M
$3.38
at Kyma API
Context window
1M
tokens
Max output
33K
tokens per reply
Hosts
1
1 listings
Released
-
date not published

Pricing

List prices at Kyma API, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.675$3.38
Cached input
80% off input
$0.135

What it costs a month

At Kyma API list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$407
  • RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$316
  • Coding agent1,000 requests a day, 40K in / 2K out, cached$559
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$2,430
Use your own numbers

Where to run it

1 host serve Nemotron 3 Ultra 550B. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
nemotron-3-ultra-550b
$0.675 / $3.38

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
Text
Produces
Text
Features
Prompt caching
Open weights
License: NVIDIA Open Model License
Runtimes: TGI, vLLM

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

No host sells it for less than Kyma API.

Smaller NVIDIA models

No cheaper NVIDIA model keeps its key capabilities.

Other frontier labs

  • Grok Build 0.1
    xAI · on xAI
    $1.00 / $2.00
    Has every key capability
    7% less
  • Gemini 3.5 Flash Lite
    Google · on Google AI Studio
    $0.30 / $2.50
    Has every key capability
    37% less
  • GPT-5.4 Nano
    OpenAI · on OpenAI
    $0.20 / $1.25
    Has every key capability
    66% less
  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    85% less

Open-weight picks

  • GLM 5.3
    Zhipu AI · on CLF AI Gateway
    $0.56 / $1.76
    Has every key capability
    36% less
  • DeepSeek V4 Pro
    DeepSeek · on CLF AI Gateway
    $0.528 / $1.58
    Has every key capability
    41% less
  • Kimi K2.7 Code
    Moonshot AI · on CLF AI Gateway
    $0.38 / $1.60
    Has every key capability
    49% less
  • Inkling Small
    Independent · on DeepInfra
    $0.45 / $1.20
    Has every key capability
    53% less

Call it

Using Kyma API's API.

Python
from litellm import completion

response = completion(
    model="kyma/nemotron-3-ultra-550b",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Nemotron 3 Ultra 550B cost?

On Kyma API, Nemotron 3 Ultra 550B costs $0.675 per million input tokens and $3.38 per million output tokens. Cached input costs $0.135 per million tokens.

Which APIs offer Nemotron 3 Ultra 550B?

Nemotron 3 Ultra 550B runs on 1 API: Kyma API.

What is Nemotron 3 Ultra 550B's context window?

1M tokens, with up to 33K tokens in one reply.

Is Nemotron 3 Ultra 550B being retired?

No retirement date is scheduled for it on any API we track.

What are cheaper alternatives to Nemotron 3 Ultra 550B?

Models that cost less and keep most of what it can do: Grok Build 0.1 on xAI (7% less), Gemini 3.5 Flash Lite on Google AI Studio (37% less) and GLM 5.3 on CLF AI Gateway (36% less). Savings are on a mix of 3 input tokens to 1 output token.

What is Nemotron 3 Ultra 550B's API model id?

Use nemotron-3-ultra-550b with Kyma API.

Sources

Raw catalog entry for nemotron-3-ultra-550b on Kyma API
{
  "max_input_tokens": 1000000,
  "max_output_tokens": 32768,
  "input_cost_per_token": 6.75e-7,
  "cache_read_input_token_cost": 1.35e-7,
  "output_cost_per_token": 0.000003375,
  "litellm_provider": "kyma",
  "mode": "chat",
  "supports_function_calling": true,
  "supports_vision": false,
  "supports_prompt_caching": true,
  "supports_reasoning": true,
  "supports_response_schema": true,
  "supports_pdf_input": false
}