Nemotron-3.5-Embed-1B-BF16 API

What is Nemotron-3.5-Embed-1B-BF16?

Use the public identifier di-nvidianemotron-35-embed-1b-bf16-2e7548780d for Nemotron-3.5-Embed-1B-BF16 in ServerNet. The gateway uses this identifier to select the configured route; the developer is nvidia.

Where to use Nemotron-3.5-Embed-1B-BF16

• Index document chunks with Nemotron-3.5-Embed-1B-BF16 for semantic search.
• Embed the search query with the same Nemotron-3.5-Embed-1B-BF16 model used to build your index.
• Evaluate retrieval quality before combining the results with a text-generation model.

Build your integration

Send a string or a list of strings to /v1/embeddings. Store the returned vectors in your search index. When switching models, rebuild the index and check vector dimensions instead of mixing incompatible vectors.

Inputs and service limits

The route accepts plain text and batches of up to 128 texts. Token ID arrays, media inputs and unsupported output formats are refused. Choose chunk sizes within the model’s configured input limit.

A request cost example

For 1,000 input tokens and 0 output tokens, without a cache discount. This is a catalogue estimate, not a fixed request fee. Reported consumption and the price frozen for your request determine the usage amount.

Estimated usage, Toman

6

Connect to Nemotron-3.5-Embed-1B-BF16 API

Enable ai:embedding on your key and ai:models:read to check availability. Set SERVERNET_API_KEY in your server environment. Change the idempotency key for each new request; keep it unchanged only for an identical retry.

This example documents the request format. Send it only after this model appears in GET /v1/models.

curl https://servernet.cloud/v1/embeddings \
  -H "Authorization: Bearer $SERVERNET_API_KEY" \
  -H 'Content-Type: application/json' \
  -H 'Idempotency-Key: CHANGE_FOR_EACH_NEW_REQUEST' \
  -d '{
    "model": "di-nvidianemotron-35-embed-1b-bf16-2e7548780d",
    "input": [
        "Hello world",
        "Search documents"
    ]
}'

Use the same Idempotency-Key when retrying an identical request. Successful responses are replayed without a second charge for 24 hours. Do not automatically retry when x-should-retry:false is present. The key is shown once. Store it in a server environment variable. Keep it out of browser code, GitHub and logs.

Compare models for the same task