Pass IndexThe State of AISign in

Qwen3-Embedding-4B

The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks.

text → text · made by Alibaba

$0.01per Mtok in
Hugging Face Inference Providers · 4 sellers

Sold by 4 ways

SellerLaneRate
Hugging Face Inference Providers aggregatorstandard$0.01per Mtok inmodels.dev · read 2026-09-16
DeepInfra aggregatorstandard$0.02per Mtok inapi.deepinfra.com · read 2026-08-25
Nous Research aggregatorstandard$0.02per Mtok ininference-api.nousresearch.com · read 2026-09-12
Vercel AI Gateway aggregatorstandard$0.02per Mtok inai-gateway.vercel.sh · read 2026-08-25

Measured 3 standings

PlaceBoardMetricScore
5thof 238MTEB · English — Reranking task-type splitscoresByTaskType.Reranking (0-1)0.5076
6thof 183MTEB · English (MTEB(eng, v2))meanTask (mean score, 0-1)0.7461
6thof 86MTEB · Multilingual (MTEB(Multilingual, v2))meanTask (mean score, 0-1)0.6945

About

Qwen3-Embedding-4B — a text model from Alibaba, sold by 4 companies from $0.01 per Mtok in, placed 5th of 238 on MTEB · English — Reranking task-type split.

It takes text and returns text, with a context window of 32,768 tokens. It was published in 2025, trained on material up to December 2024. The catalogue files it under rerank. Four companies sell it. The cheapest is $0.01 per Mtok in at Hugging Face Inference Providers. It has been measured on 3 boards, and stands best at 5th of 238 on MTEB · English — Reranking task-type split.

Every current figure

Maker
Alibaba
Register
model
Takes
text
Returns
text
Context
32,768 tokens
Published
2025
Knowledge to
December 2024
Parameters
4 billion · read from its own name
Licence
not read
Sellers
4
Maker's own price
not read
Price
$0.01 per Mtok in — Hugging Face Inference Providers
Boards
3
Best place
5th of 238 — MTEB · English — Reranking task-type split

Known as 2 names

Qwen/Qwen3-Embedding-4Balibaba/qwen3-embedding-4b