Embedding Changelog

Last updated: October 21, 2026

This page tracks native embedding model additions, deprecations, and price changes over time. For the current catalog, see supported models. For help estimating cost, contact us.

October 2026

Models added

  • nvidia/nemotron-3.5-embed-1b at $0.02/1M tokens
  • nvidia/nemotron-3.5-embed-8b at $0.05/1M tokens

September 2026

Native embeddings is generally available.

Pricing changes

  • nvidia/nemotron-3-embed-1b priced at $0.02/1M tokens

August 2026

Models added

  • voyage/voyage-code-4 at $0.12/1M tokens
  • openai/text-embedding-3-large at $0.13/1M tokens
  • openai/text-embedding-3-small at $0.02/1M tokens
  • openai/text-embedding-ada-002 at $0.10/1M tokens

Models deprecated

  • cohere/embed-english-v3.0
  • google/gemini-embedding-001
  • google/text-embedding-005, support ends: April 1, 2027
  • voyage/voyage-3-large
  • voyage/voyage-3.5
  • voyage/voyage-3.5-lite

Pricing changes

  • voyage/voyage-4-nano priced at $0.016/1M tokens
  • nvidia/nemotron-3-embed-8b priced at $0.05/1M tokens

July 2026

Launched native embeddings in private beta.

Models added

  • baai/bge-m3 at $0.01/1M tokens
  • cohere/embed-english-v3.0 at $0.10/1M tokens
  • cohere/embed-v4.0 at $0.12/1M tokens
  • google/gemini-embedding-001 at $0.15/1M tokens
  • google/gemini-embedding-2 at $0.20/1M tokens
  • nvidia/nemotron-3-embed-1b
  • nvidia/nemotron-3-embed-8b
  • qwen/qwen3-embedding-0p6b at $0.01/1M tokens
  • qwen/qwen3-embedding-4b at $0.02/1M tokens
  • qwen/qwen3-embedding-8b at $0.07/1M tokens
  • voyage/voyage-4 at $0.06/1M tokens
  • voyage/voyage-4-large at $0.12/1M tokens
  • voyage/voyage-4-lite at $0.02/1M tokens
  • voyage/voyage-4-nano
  • voyage/voyage-code-3 at $0.18/1M tokens
  • zeroentropy/zembed-1 at $0.05/1M tokens