Text embeddings

Ranked by an independent measurement first, then by what it costs per 1M input tokens across sellers. Absence of a score is not a zero; absence of a price is not free.

2 undominated of 7 scored and priced

As of . 57 models on this benchmark · 20 scored · 31 priced · 13 scored but unpriced · 24 priced but unrated.

The frontier

$0.0030$0.010$0.030$0.10060657075Price per 1M input tokensMTEB(eng, v2) mean
Each dot is a model that is both independently scored and priced. The staircase is the frontier: nothing on the page beats a step on score and undercuts it on price.

MTEB(eng, v2)

Sorted by the MTEB(eng, v2) mean over 41 tasks. A model with fewer than 41 tasks has no mean and is listed as partial with its count.

RankModelMTEB(eng, v2) meanPriceSellersSpreadVerdict
QZhou-Embedding Kingsoft OW75.97 no traceable priceunpriced
Qwen3-Embedding-8B Alibaba Qwen OW75.23 $0.010 DeepInfra310×undominated
Jasper-Token-Compression-600M infgrad OW74.75 no traceable priceunpriced
Qwen3-Embedding-4B Alibaba Qwen OW74.61 $0.020 DeepInfra2dominated by Qwen3-Embedding-8B
Seed1.6-embedding ByteDance 74.07 no traceable priceunpriced
gemini-embedding-001 Google 73.30 $0.150 OpenRouter1dominated by Qwen3-Embedding-8B
jina-embeddings-v5-text-nano Jina AI OW71.11 no traceable priceunpriced
Qwen3-Embedding-0.6B Alibaba Qwen OW70.47 $0.010 DeepInfra1dominated by Qwen3-Embedding-8B
EmbeddingGemma 300M Google OW69.67 $0.0020 DeepInfra1undominated
text-embedding-005 Google 69.60 no traceable priceunpriced
text-multilingual-embedding-002 Google 67.67 no traceable priceunpriced
mxbai-embed-large-v1 Mixedbread OW66.26 no traceable priceunpriced
multilingual-e5-large-instruct Microsoft (intfloat) OW65.53 $0.010 DeepInfra1dominated by EmbeddingGemma 300M
nomic-embed-text-v2-moe Nomic OW63.64 no traceable priceunpriced
snowflake-arctic-embed-l-v2.0 Snowflake OW63.60 no traceable priceunpriced
granite-embedding-english-r2 IBM OW62.84 no traceable priceunpriced
voyage-3 Voyage AI 61.38 no traceable priceunpriced
voyage-3-lite Voyage AI 60.44 no traceable priceunpriced
all-MiniLM-L6-v2 Sentence Transformers OW59.03 $0.0050 DeepInfra2dominated by EmbeddingGemma 300M
paraphrase-multilingual-MiniLM-L12-v2 Sentence Transformers OW56.42 no traceable priceunpriced
Nemotron-3-Embed-1B NVIDIA OWpartial, 0 of 41 free OpenRouter2unrated
pplx-embed-v1-0.6b Perplexity partial, 0 of 41 $0.0040 OpenRouter1unrated
bge-base-en-v1.5 BAAI OWpartial, 33 of 41 $0.0050 DeepInfra2unrated
e5-base-v2 Microsoft (intfloat) OWpartial, 33 of 41 $0.0050 DeepInfra2unrated
all-MiniLM-L12-v2 Sentence Transformers OWpartial, 33 of 41 $0.0050 DeepInfra2unrated
all-mpnet-base-v2 Sentence Transformers OWpartial, 33 of 41 $0.0050 DeepInfra2unrated
multi-qa-mpnet-base-dot-v1 Sentence Transformers OWpartial, 0 of 41 $0.0050 DeepInfra2unrated
gte-base Alibaba DAMO (thenlper) OWpartial, 21 of 41 $0.0050 DeepInfra2unrated
bge-large-en-v1.5 BAAI OWpartial, 33 of 41 $0.010 DeepInfra2unrated
bge-m3 BAAI OWpartial, 27 of 41 $0.010 DeepInfra2unrated
e5-large-v2 Microsoft (intfloat) OWpartial, 33 of 41 $0.010 DeepInfra2unrated
multilingual-e5-large Microsoft (intfloat) OWpartial, 23 of 41 $0.010 DeepInfra2unrated
gte-large Alibaba DAMO (thenlper) OWpartial, 21 of 41 $0.010 DeepInfra2unrated
text-embedding-3-small OpenAI partial, 6 of 41 $0.020 OpenRouter1unrated
voyage-4-lite Voyage AI partial, 0 of 41 $0.020 OpenRouter1unrated
pplx-embed-v1-4b Perplexity partial, 0 of 41 $0.030 OpenRouter1unrated
Nemotron-3-Embed-8B NVIDIA OWpartial, 0 of 41 $0.035 DeepInfra1unrated
voyage-4 Voyage AI partial, 0 of 41 $0.060 OpenRouter1unrated
mistral-embed Mistral partial, 0 of 41 $0.100 OpenRouter1unrated
text-embedding-ada-002 OpenAI partial, 0 of 41 $0.100 OpenRouter1unrated
voyage-4-large Voyage AI partial, 0 of 41 $0.120 OpenRouter1unrated
text-embedding-3-large OpenAI partial, 10 of 41 $0.130 OpenRouter1unrated
codestral-embed Mistral partial, 0 of 41 $0.150 OpenRouter1unrated
gemini-embedding-2 Google partial, 0 of 41 $0.200 OpenRouter1unrated
gte-modernbert-base Alibaba NLP OWpartial, 27 of 41 no traceable priceno data
gte-Qwen2-7B-instruct Alibaba NLP OWpartial, 33 of 41 no traceable priceno data
Titan Text Embeddings V2 Amazon partial, 27 of 41 no traceable priceno data
embed-english-v3.0 Cohere partial, 21 of 41 no traceable priceno data
embed-multilingual-v3.0 Cohere partial, 21 of 41 no traceable priceno data
embed-v4.0 Cohere partial, 0 of 41 no traceable priceno data
jina-embeddings-v3 Jina AI OWpartial, 30 of 41 no traceable priceno data
jina-embeddings-v4 Jina AI OWpartial, 19 of 41 no traceable priceno data
jina-embeddings-v5-text-small Jina AI OWpartial, 40 of 41 no traceable priceno data
nomic-embed-text-v1.5 Nomic OWpartial, 33 of 41 no traceable priceno data
voyage-3-large Voyage AI partial, 3 of 41 no traceable priceno data
voyage-3.5 Voyage AI partial, 22 of 41 no traceable priceno data
voyage-code-3 Voyage AI partial, 29 of 41 no traceable priceno data

Worth shopping

The same model at more than one seller, dearest divided by cheapest, over distinct sellers at the reference unit. A vendor's own tiers are not competition.

ModelSpreadSellersCheapest
Qwen3-Embedding-8B10×3$0.010 DeepInfra

How this page is built

  • Quality. The mean over the 41 tasks of MTEB(eng, v2), computed here from the CC0 per-task result files. Scores reach that repository as pull requests, usually from the vendor.
  • Price. Vendor pages read by hand, plus the public price feeds of OpenRouter and DeepInfra. Every figure carries the page it came from and the day it was read. Community aggregations are used to find a vendor page, never as the number.
  • Normalisation. One reference unit, per 1M input tokens. Conversions are printed next to each figure on the model page. A price stated under conditions that cannot be converted is shown and excluded from the ranking rather than guessed at.
  • Verdict. Undominated means nothing here is both better and cheaper. Computed over models that are both scored and priced; the rest are named by their state.

MTEB results, https://github.com/embeddings-benchmark/results, released under CC0-1.0; MTEB(eng, v2) means computed from the per-task files as described in data/modalities/embeddings/benchmark.json.

Questions this page answers

What does the score on this page measure?

It is the mean over the 41 tasks of MTEB(eng, v2), computed from the CC0 results repository rather than copied from a leaderboard. MTEB publishes no confidence interval, so none is shown, and a model missing any of the 41 tasks gets no mean at all.

What is a price per 1M input tokens here?

Every seller states a price in its own unit. Each figure is normalised to one reference unit, per 1M input tokens, with the rule printed beside it on the model page, and nothing is rescaled across resolutions or quality tiers. A figure that cannot be converted without inventing something is shown, not ranked.

Why do so many models have no price?

13 of these models are scored but have no price that traces to a page we have read. Missing is not free, and an unpriced model is never called cheap. Most vendors in this market publish prices as JavaScript pages or not at all.

What does undominated mean?

Nothing else on this page is both better on the score and cheaper on the price. Dominated rows name the model that beats them on both. A model that is unrated or unpriced is neither: it is not ranked at all.