Embeddings: compare 11 models
Vectors for search and RAG. Prices are the providers' own, in this category's unit.
| # | Model | Price | Context | Released | Price change | Compare |
|---|---|---|---|---|---|---|
| 1 |
text-embedding-3-large
OpenAI
|
$0.13 / 1M tokens | 8k | January 2024 | ||
| 2 |
Gemini Embedding 2
Google
|
$0.20 / 1M tokens | 8k | April 2026 | ||
| 3 |
Embed v5.0 Pro
Cohere
|
No data | 128k | No data | ||
| 4 |
text-embedding-3-small
OpenAI
|
$0.02 / 1M tokens | 8k | January 2024 | ||
| 5 |
Codestral Embed
Mistral AI
|
No data | 8k | May 2025 | ||
| 6 |
Mistral Embed
Mistral AI
|
No data | 8k | December 2023 | ||
| 7 |
Embed v5.0 Fast
Cohere
|
No data | 128k | No data | ||
| 8 |
Qwen3.7 Text Embedding
Alibaba Cloud (Qwen)
|
No data | No data | No data | ||
| 9 |
text-embedding-ada-002
Legacy
OpenAI
|
$0.10 / 1M tokens | 8k | December 2022 | ||
| 10 |
Embed Multilingual v3.0
Legacy
Cohere
|
No data | 512 | November 2023 | ||
| 11 |
Embed v4.0
Legacy
Cohere
|
No data | 128k | April 2025 |
Frequently asked questions
What is the cheapest model for embeddings?
Among current models with a published price, text-embedding-3-small from OpenAI is the lowest at $0.02 / 1M tokens (checked Oct 22, 2026). Prices change, so confirm on the provider's page.
How many embeddings models are available?
8 current models from 5 providers: OpenAI, Google, Cohere, Mistral AI, Alibaba Cloud (Qwen).