text-embedding-3-large
OpenAIEmbedding Models
91.0
Performance
★ 4.5
Rating
890
Reviews
EmbeddingProprietary
About
OpenAI's most capable embedding model producing 3,072-dimensional vectors for search, RAG, classification, and clustering.
Strengths
Highest quality embeddings from OpenAI with 3,072 dimensions. Strong on retrieval, classification, and clustering tasks. Supports native dimension reduction (256-3072) for cost-quality tradeoffs. Matryoshka representation learning for flexible dimensions.
Specifications
- Context window
- 8,191
- Parameters
- —
Pricing
- Input cost
- $0.13 / 1M
- Output cost
- —
$0.13 per 1M tokens
Available On
OpenAI APIAzure OpenAI
Features
text embeddingdimension reductionbatch processing
Market Sentiment
Polymarket prediction market odds for "Best AI Model" — showing how the market values each provider over time.
Loading sentiment data...
Performance Trend
Benchmark score trends over time for the top 5 benchmarks.
Loading history...
Benchmarks
Scores from various benchmark tests; higher is better.
| Test | Score | Percentile | Source |
|---|---|---|---|
| MTEB | 64.6 | p90 | seed |