Skip to content
Tier B — Production
Runs in:US
OpenAI

OpenAI text-embedding-3-large

Tier B — Production

Tokonomix Editorial Team·Reviewed by Mes Kalkan·
Section 01

Pricing

What you pay per million tokens when you use this model on Tokonomix, plus an estimate for a typical conversation.

💰
API rates — OpenAI text-embedding-3-large
$0.1700 per 1M input tokens
— per 1M output tokens
≈ $0.0001 per typical conversation (800 tokens)
Input vs output price (per 1M tokens)
per 1M input tokens$0.1700
per 1M output tokens—
Section 02

Availability

Availability

No measurements yet

We haven't recorded enough API calls to show availability stats for this model. Data appears once the model starts receiving live traffic.

Section 03

Tokonomix benchmark verdicts

● 2026-06-21

First benchmark establishes baseline for text-embedding-3-large

OpenAI's text-embedding-3-large enters benchmarking with strong performance across multiple evaluation domains. The model demonstrates particular strength in retrieval tasks, achieving 54.90 on NDCG@10 and 49.40 on the MIRACL benchmark, indicating robust multilingual retrieval capabilities. Classification performance stands at 71.15, while clustering reaches 47.80, showing balanced competency across different embedding use cases. The model produces 3072-dimensional embeddings with a context window of 8191 tokens, providing substantial capacity for processing longer documents. Reranking capabilities score at 59.36, positioning this as a versatile embedding model suitable for various semantic search and information retrieval applications. The STS (Semantic Textual Similarity) score of 53.26 reflects solid performance in understanding nuanced semantic relationships. As a large-scale embedding model, it appears designed for production environments requiring high-quality vector representations across diverse languages and tasks. Users should note this baseline establishes the expected performance envelope, with future benchmarks tracking consistency and any performance shifts over time.

Quality

—

Latency p50

—

Test runs

0

✓ Strong retrieval performance established✓ Multilingual capabilities confirmed✓ Large 3072-dimensional embeddings✓ 8191 token context window
Section 04

Full model profile

text-embedding-3-large: OpenAI's high-fidelity embedding model, dimensions and all

text-embedding-3-large is OpenAI's higher-quality embedding model, released January 25, 2024 alongside the smaller text-embedding-3-small. An embedding model does not generate text — it converts a piece of text into a fixed-length vector of numbers that captures its meaning, so that two pieces of text with similar meaning end up as vectors that sit close together in that numeric space. That property is what makes embeddings useful for semantic search, clustering, deduplication, recommendation, and retrieval-augmented generation (RAG), where a system needs to find the passages most relevant to a query before handing them to a language model.

What "large" buys over "small"

By default, text-embedding-3-large produces a 3,072-dimension vector, versus 1,536 for text-embedding-3-small. On OpenAI's own MTEB benchmark, a standard suite for evaluating embedding quality across retrieval, classification, and clustering tasks, large scores 64.6 against small's 62.3 — a real but not enormous quality gap. The larger model captures more nuance per embedding, which tends to matter most on harder retrieval tasks: distinguishing between closely related documents, working across less common domains or languages, or ranking many similar candidates against each other.

The dimensions parameter

Both models support a dimensions parameter that shortens the output vector by truncating it, without needing to re-embed the text through a different model. According to OpenAI, a text-embedding-3-large vector shortened to 256 dimensions still outperforms the older, unshortened text-embedding-ada-002 model at its native 1,536 dimensions. This matters operationally: a shorter vector takes less storage in a vector database and is cheaper and faster to compare at query time, so the dimensions parameter is a real lever for trading a small amount of retrieval quality for meaningfully lower index size and search latency — without switching models or re-architecting the pipeline.

Under the hood

Maximum input length is 8,192 tokens per request. Output is embedding-only — there is no text generation, no chat interface, no tool use. The model is text-in, vector-out, and nothing else, which keeps its integration surface small: a single API call takes a string and returns a fixed-size array of floats.

When to pick large over small

Reach for text-embedding-3-large when retrieval quality is the bottleneck: a RAG system missing relevant passages, a search feature that returns near-misses instead of the right result, or a clustering task where fine-grained distinctions between similar items matter. It is also the more defensible default when working in less common languages or specialized domains, where a smaller embedding space has less room to represent nuance.

For very high-volume indexing — millions of documents, or a workload where query latency and storage cost dominate the design — text-embedding-3-small, or a truncated large embedding via the dimensions parameter, is usually the more practical starting point. The MTEB gap between the two models is real but modest, so it is worth testing both against a representative sample of your own data rather than defaulting to large purely on the assumption that bigger is better.

Alternatives worth comparing

text-embedding-ada-002 remains available as OpenAI's previous-generation embedding model, but it is dominated on quality by both current models even at a fraction of their dimensionality, so there is little reason to choose it for new work. Outside OpenAI's own lineup, several other providers ship general-purpose text embedding models with comparable positioning; the right choice usually comes down to testing retrieval quality on your own corpus rather than benchmark scores alone, since MTEB performance does not always transfer cleanly to a specific domain.

Last automated test
Jun 21, 2026 · 04:48 UTC · Benchmark
P50 latency
—
P95 latency
—
Errors
1 / 3 runs
Last reviewed by Tokonomix Team·September 14, 2026