> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinecone.io/llms.txt
> Use this file to discover all available pages before exploring further.
# List available models
> List the embedding and reranking models hosted by Pinecone.
You can use hosted models as an integrated part of Pinecone operations or for standalone embedding and reranking. For more details, see [Vector embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding) and [Rerank results](https://docs.pinecone.io/guides/search/rerank-results).
```python Python theme={null}
from pinecone import Pinecone
pc = Pinecone(api_key="YOUR_API_KEY")
models = pc.inference.list_models()
print(models)
```
```javascript JavaScript theme={null}
const pc = new Pinecone({ apiKey: 'YOUR_API_KEY' });
const models = await pc.inference.listModels();
console.log(models);
```
```java Java theme={null}
import io.pinecone.clients.Inference;
import io.pinecone.clients.Pinecone;
import org.openapitools.inference.client.ApiException;
import org.openapitools.inference.client.model.ModelInfo;
import org.openapitools.inference.client.model.ModelInfoList;
public class ListModels {
public static void main(String[] args) throws ApiException {
Pinecone pinecone = new Pinecone.Builder("YOUR_API_KEY").build();
Inference inference = pinecone.getInferenceClient();
// List all models
ModelInfoList models = inference.listModels();
System.out.println(models);
// List by model type ("embed" or "rerank")
ModelInfoList modelsByModelType = inference.listModels("rerank");
System.out.println(modelsByModelType);
// List by model type ("embed" or "rerank") and vector type ("dense" or "sparse")
ModelInfoList modelsByModelTypeAndVectorType = inference.listModels("embed", "dense");
System.out.println(modelsByModelTypeAndVectorType);
}
}
```
```go Go theme={null}
package main
import (
"context"
"encoding/json"
"fmt"
"log"
"github.com/pinecone-io/go-pinecone/v4/pinecone"
)
func prettifyStruct(obj interface{}) string {
bytes, _ := json.MarshalIndent(obj, "", " ")
return string(bytes)
}
func main() {
ctx := context.Background()
pc, err := pinecone.NewClient(pinecone.NewClientParams{
ApiKey: "YOUR_API_KEY",
})
if err != nil {
log.Fatalf("Failed to create Client: %v", err)
}
embed := "embed"
rerank := "rerank"
embedModels, err := pc.Inference.ListModels(ctx, &pinecone.ListModelsParams{
Type: &embed,
})
if err != nil {
log.Fatalf("Failed to list embedding models: %v", err)
}
fmt.Printf(prettifyStruct(embedModels))
rerankModels, err := pc.Inference.ListModels(ctx, &pinecone.ListModelsParams{
Type: &rerank,
})
if err != nil {
log.Fatalf("Failed to list reranking models: %v", err)
}
fmt.Printf(prettifyStruct(rerankModels))
}
```
```csharp C# theme={null}
using Pinecone;
using Pinecone.Inference;
var pinecone = new PineconeClient("YOUR_API_KEY");
var models = await pinecone.Inference.Models.ListAsync(new ListModelsRequest());
Console.WriteLine(models);
```
```bash curl theme={null}
PINECONE_API_KEY="YOUR_API_KEY"
curl "https://api.pinecone.io/models" \
-H "Api-Key: $PINECONE_API_KEY" \
-H "X-Pinecone-Api-Version: 2025-04"
```
```python Python theme={null}
[{
"model": "llama-text-embed-v2",
"short_description": "A high performance dense embedding model optimized for multilingual and cross-lingual text question-answering retrieval with support for long documents (up to 2048 tokens) and dynamic embedding size (Matryoshka Embeddings).",
"type": "embed",
"supported_parameters": [
{
"parameter": "input_type",
"type": "one_of",
"value_type": "string",
"required": true,
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"default": "END",
"allowed_values": [
"END",
"NONE",
"START"
]
},
{
"parameter": "dimension",
"type": "one_of",
"value_type": "integer",
"required": false,
"default": 1024,
"allowed_values": [
384,
512,
768,
1024,
2048
]
}
],
"vector_type": "dense",
"default_dimension": 1024,
"modality": "text",
"max_sequence_length": 2048,
"max_batch_size": 96,
"provider_name": "NVIDIA",
"supported_metrics": [
"cosine",
"dotproduct"
],
"supported_dimensions": [
384,
512,
768,
1024,
2048
]
}, {
"model": "multilingual-e5-large",
"short_description": "A high-performance dense embedding model trained on a mixture of multilingual datasets. It works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
"type": "embed",
"supported_parameters": [
{
"parameter": "input_type",
"type": "one_of",
"value_type": "string",
"required": true,
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"default": "END",
"allowed_values": [
"END",
"NONE"
]
}
],
"vector_type": "dense",
"default_dimension": 1024,
"modality": "text",
"max_sequence_length": 507,
"max_batch_size": 96,
"provider_name": "Microsoft",
"supported_metrics": [
"cosine",
"euclidean"
],
"supported_dimensions": [
1024
]
}, {
"model": "pinecone-sparse-english-v0",
"short_description": "A sparse embedding model for converting text to sparse vectors for keyword or hybrid semantic/keyword search. Built on the innovations of the DeepImpact architecture.",
"type": "embed",
"supported_parameters": [
{
"parameter": "input_type",
"type": "one_of",
"value_type": "string",
"required": true,
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"default": "END",
"allowed_values": [
"END",
"NONE"
]
},
{
"parameter": "return_tokens",
"type": "any",
"value_type": "boolean",
"required": false,
"default": false
}
],
"vector_type": "sparse",
"modality": "text",
"max_sequence_length": 512,
"max_batch_size": 96,
"provider_name": "Pinecone",
"supported_metrics": [
"dotproduct"
]
}, {
"model": "bge-reranker-v2-m3",
"short_description": "A high-performance, multilingual reranking model that works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
"type": "rerank",
"supported_parameters": [
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"default": "NONE",
"allowed_values": [
"END",
"NONE"
]
}
],
"modality": "text",
"max_sequence_length": 1024,
"max_batch_size": 100,
"provider_name": "BAAI",
"supported_metrics": []
}, {
"model": "cohere-rerank-3.5",
"short_description": "Cohere's leading reranking model, balancing performance and latency for a wide range of enterprise search applications.",
"type": "rerank",
"supported_parameters": [
{
"parameter": "max_chunks_per_doc",
"type": "numeric_range",
"value_type": "integer",
"required": false,
"default": 3072,
"min": 1.0,
"max": 3072.0
}
],
"modality": "text",
"max_sequence_length": 40000,
"max_batch_size": 200,
"provider_name": "Cohere",
"supported_metrics": []
}, {
"model": "pinecone-rerank-v0",
"short_description": "A state of the art reranking model that out-performs competitors on widely accepted benchmarks. It can handle chunks up to 512 tokens (1-2 paragraphs)",
"type": "rerank",
"supported_parameters": [
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"default": "END",
"allowed_values": [
"END",
"NONE"
]
}
],
"modality": "text",
"max_sequence_length": 512,
"max_batch_size": 100,
"provider_name": "Pinecone",
"supported_metrics": []
}]
```
```javascript JavaScript theme={null}
{
models: [
{
model: 'llama-text-embed-v2',
shortDescription: 'A high performance dense embedding model optimized for multilingual and cross-lingual text question-answering retrieval with support for long documents (up to 2048 tokens) and dynamic embedding size (Matryoshka Embeddings).',
type: 'embed',
vectorType: 'dense',
defaultDimension: 1024,
modality: 'text',
maxSequenceLength: 2048,
maxBatchSize: 96,
providerName: 'NVIDIA',
supportedDimensions: [Array],
supportedMetrics: [Array],
supportedParameters: [Array]
},
{
model: 'multilingual-e5-large',
shortDescription: 'A high-performance dense embedding model trained on a mixture of multilingual datasets. It works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)',
type: 'embed',
vectorType: 'dense',
defaultDimension: 1024,
modality: 'text',
maxSequenceLength: 507,
maxBatchSize: 96,
providerName: 'Microsoft',
supportedDimensions: [Array],
supportedMetrics: [Array],
supportedParameters: [Array]
},
{
model: 'pinecone-sparse-english-v0',
shortDescription: 'A sparse embedding model for converting text to sparse vectors for keyword or hybrid semantic/keyword search. Built on the innovations of the DeepImpact architecture.',
type: 'embed',
vectorType: 'sparse',
defaultDimension: undefined,
modality: 'text',
maxSequenceLength: 512,
maxBatchSize: 96,
providerName: 'Pinecone',
supportedDimensions: undefined,
supportedMetrics: [Array],
supportedParameters: [Array]
},
{
model: 'bge-reranker-v2-m3',
shortDescription: 'A high-performance, multilingual reranking model that works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)',
type: 'rerank',
vectorType: undefined,
defaultDimension: undefined,
modality: 'text',
maxSequenceLength: 1024,
maxBatchSize: 100,
providerName: 'BAAI',
supportedDimensions: undefined,
supportedMetrics: undefined,
supportedParameters: [Array]
},
{
model: 'cohere-rerank-3.5',
shortDescription: "Cohere's leading reranking model, balancing performance and latency for a wide range of enterprise search applications.",
type: 'rerank',
vectorType: undefined,
defaultDimension: undefined,
modality: 'text',
maxSequenceLength: 40000,
maxBatchSize: 200,
providerName: 'Cohere',
supportedDimensions: undefined,
supportedMetrics: undefined,
supportedParameters: [Array]
},
{
model: 'pinecone-rerank-v0',
shortDescription: 'A state of the art reranking model that out-performs competitors on widely accepted benchmarks. It can handle chunks up to 512 tokens (1-2 paragraphs)',
type: 'rerank',
vectorType: undefined,
defaultDimension: undefined,
modality: 'text',
maxSequenceLength: 512,
maxBatchSize: 100,
providerName: 'Pinecone',
supportedDimensions: undefined,
supportedMetrics: undefined,
supportedParameters: [Array]
}
]
}
```
```java Java theme={null}
class ModelInfoList {
models: [class ModelInfo {
model: llama-text-embed-v2
shortDescription: A high performance dense embedding model optimized for multilingual and cross-lingual text question-answering retrieval with support for long documents (up to 2048 tokens) and dynamic embedding size (Matryoshka Embeddings).
type: embed
vectorType: dense
defaultDimension: 1024
modality: text
maxSequenceLength: 2048
maxBatchSize: 96
providerName: NVIDIA
supportedDimensions: [384, 512, 768, 1024, 2048]
supportedMetrics: [cosine, dotproduct]
supportedParameters: [class ModelInfoSupportedParameter {
parameter: input_type
type: one_of
valueType: string
required: true
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: query
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: passage
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: null
additionalProperties: null
}, class ModelInfoSupportedParameter {
parameter: truncate
type: one_of
valueType: string
required: false
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: END
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: NONE
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: START
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: END
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}, class ModelInfoSupportedParameter {
parameter: dimension
type: one_of
valueType: integer
required: false
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: 384
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: 512
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: 768
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: 1024
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: 2048
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: 1024
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}]
additionalProperties: null
}, class ModelInfo {
model: multilingual-e5-large
shortDescription: A high-performance dense embedding model trained on a mixture of multilingual datasets. It works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)
type: embed
vectorType: dense
defaultDimension: 1024
modality: text
maxSequenceLength: 507
maxBatchSize: 96
providerName: Microsoft
supportedDimensions: [1024]
supportedMetrics: [cosine, euclidean]
supportedParameters: [class ModelInfoSupportedParameter {
parameter: input_type
type: one_of
valueType: string
required: true
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: query
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: passage
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: null
additionalProperties: null
}, class ModelInfoSupportedParameter {
parameter: truncate
type: one_of
valueType: string
required: false
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: END
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: NONE
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: END
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}]
additionalProperties: null
}, class ModelInfo {
model: pinecone-sparse-english-v0
shortDescription: A sparse embedding model for converting text to sparse vectors for keyword or hybrid semantic/keyword search. Built on the innovations of the DeepImpact architecture.
type: embed
vectorType: sparse
defaultDimension: null
modality: text
maxSequenceLength: 512
maxBatchSize: 96
providerName: Pinecone
supportedDimensions: null
supportedMetrics: [dotproduct]
supportedParameters: [class ModelInfoSupportedParameter {
parameter: input_type
type: one_of
valueType: string
required: true
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: query
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: passage
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: null
additionalProperties: null
}, class ModelInfoSupportedParameter {
parameter: truncate
type: one_of
valueType: string
required: false
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: END
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: NONE
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: END
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}, class ModelInfoSupportedParameter {
parameter: return_tokens
type: any
valueType: boolean
required: false
allowedValues: null
min: null
max: null
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: false
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}, class ModelInfoSupportedParameter {
parameter: max_tokens_per_sequence
type: one_of
valueType: integer
required: false
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: 512
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: 2048
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: 512
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}]
additionalProperties: null
}, class ModelInfo {
model: bge-reranker-v2-m3
shortDescription: A high-performance, multilingual reranking model that works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)
type: rerank
vectorType: null
defaultDimension: null
modality: text
maxSequenceLength: 1024
maxBatchSize: 100
providerName: BAAI
supportedDimensions: null
supportedMetrics: null
supportedParameters: [class ModelInfoSupportedParameter {
parameter: truncate
type: one_of
valueType: string
required: false
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: END
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: NONE
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: NONE
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}]
additionalProperties: null
}, class ModelInfo {
model: cohere-rerank-3.5
shortDescription: Cohere's leading reranking model, balancing performance and latency for a wide range of enterprise search applications.
type: rerank
vectorType: null
defaultDimension: null
modality: text
maxSequenceLength: 40000
maxBatchSize: 200
providerName: Cohere
supportedDimensions: null
supportedMetrics: null
supportedParameters: [class ModelInfoSupportedParameter {
parameter: max_chunks_per_doc
type: numeric_range
valueType: integer
required: false
allowedValues: null
min: 1
max: 3072
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: 3072
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}]
additionalProperties: null
}, class ModelInfo {
model: pinecone-rerank-v0
shortDescription: A state of the art reranking model that out-performs competitors on widely accepted benchmarks. It can handle chunks up to 512 tokens (1-2 paragraphs)
type: rerank
vectorType: null
defaultDimension: null
modality: text
maxSequenceLength: 512
maxBatchSize: 100
providerName: Pinecone
supportedDimensions: null
supportedMetrics: null
supportedParameters: [class ModelInfoSupportedParameter {
parameter: truncate
type: one_of
valueType: string
required: false
allowedValues: [class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: END
isNullable: false
schemaType: anyOf
}, class class org.openapitools.inference.client.model.ModelInfoSupportedParameterAllowedValuesInner {
instance: NONE
isNullable: false
schemaType: anyOf
}]
min: null
max: null
_default: class class org.openapitools.inference.client.model.ModelInfoSupportedParameterDefault {
instance: END
isNullable: false
schemaType: anyOf
}
additionalProperties: null
}]
additionalProperties: null
}]
additionalProperties: null
}
```
```go Go theme={null}
{
"models": [
{
"default_dimension": 1024,
"max_batch_size": 96,
"max_sequence_length": 2048,
"modality": "text",
"model": "llama-text-embed-v2",
"provider_name": "NVIDIA",
"short_description": "A high performance dense embedding model optimized for multilingual and cross-lingual text question-answering retrieval with support for long documents (up to 2048 tokens) and dynamic embedding size (Matryoshka Embeddings).",
"supported_dimensions": [
384,
512,
768,
1024,
2048
],
"supported_metrics": [
"cosine",
"dotproduct"
],
"supported_parameters": [
{
"allowed_values": [
{
"StringValue": "query",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "passage",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
}
],
"parameter": "input_type",
"required": true,
"type": "one_of",
"value_type": "string"
},
{
"allowed_values": [
{
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "NONE",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "START",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
}
],
"default": {
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
"parameter": "truncate",
"required": false,
"type": "one_of",
"value_type": "string"
},
{
"allowed_values": [
{
"StringValue": null,
"IntValue": 384,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": null,
"IntValue": 512,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": null,
"IntValue": 768,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": null,
"IntValue": 1024,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": null,
"IntValue": 2048,
"FloatValue": null,
"BoolValue": null
}
],
"default": {
"StringValue": null,
"IntValue": 1024,
"FloatValue": null,
"BoolValue": null
},
"parameter": "dimension",
"required": false,
"type": "one_of",
"value_type": "integer"
}
],
"type": "embed",
"vector_type": "dense"
},
{
"default_dimension": 1024,
"max_batch_size": 96,
"max_sequence_length": 507,
"modality": "text",
"model": "multilingual-e5-large",
"provider_name": "Microsoft",
"short_description": "A high-performance dense embedding model trained on a mixture of multilingual datasets. It works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
"supported_dimensions": [
1024
],
"supported_metrics": [
"cosine",
"euclidean"
],
"supported_parameters": [
{
"allowed_values": [
{
"StringValue": "query",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "passage",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
}
],
"parameter": "input_type",
"required": true,
"type": "one_of",
"value_type": "string"
},
{
"allowed_values": [
{
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "NONE",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
}
],
"default": {
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
"parameter": "truncate",
"required": false,
"type": "one_of",
"value_type": "string"
}
],
"type": "embed",
"vector_type": "dense"
},
{
"max_batch_size": 96,
"max_sequence_length": 512,
"modality": "text",
"model": "pinecone-sparse-english-v0",
"provider_name": "Pinecone",
"short_description": "A sparse embedding model for converting text to sparse vectors for keyword or hybrid semantic/keyword search. Built on the innovations of the DeepImpact architecture.",
"supported_metrics": [
"dotproduct"
],
"supported_parameters": [
{
"allowed_values": [
{
"StringValue": "query",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "passage",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
}
],
"parameter": "input_type",
"required": true,
"type": "one_of",
"value_type": "string"
},
{
"allowed_values": [
{
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "NONE",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
}
],
"default": {
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
"parameter": "truncate",
"required": false,
"type": "one_of",
"value_type": "string"
},
{
"default": {
"StringValue": null,
"IntValue": null,
"FloatValue": null,
"BoolValue": false
},
"parameter": "return_tokens",
"required": false,
"type": "any",
"value_type": "boolean"
},
{
"allowed_values": [
{
"StringValue": null,
"IntValue": 512,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": null,
"IntValue": 2048,
"FloatValue": null,
"BoolValue": null
}
],
"default": {
"StringValue": null,
"IntValue": 512,
"FloatValue": null,
"BoolValue": null
},
"parameter": "max_tokens_per_sequence",
"required": false,
"type": "one_of",
"value_type": "integer"
}
],
"type": "embed",
"vector_type": "sparse"
}
]
}{
"models": [
{
"max_batch_size": 100,
"max_sequence_length": 1024,
"modality": "text",
"model": "bge-reranker-v2-m3",
"provider_name": "BAAI",
"short_description": "A high-performance, multilingual reranking model that works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
"supported_parameters": [
{
"allowed_values": [
{
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "NONE",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
}
],
"default": {
"StringValue": "NONE",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
"parameter": "truncate",
"required": false,
"type": "one_of",
"value_type": "string"
}
],
"type": "rerank"
},
{
"max_batch_size": 200,
"max_sequence_length": 40000,
"modality": "text",
"model": "cohere-rerank-3.5",
"provider_name": "Cohere",
"short_description": "Cohere's leading reranking model, balancing performance and latency for a wide range of enterprise search applications.",
"supported_parameters": [
{
"default": {
"StringValue": null,
"IntValue": 3072,
"FloatValue": null,
"BoolValue": null
},
"max": 3072,
"min": 1,
"parameter": "max_chunks_per_doc",
"required": false,
"type": "numeric_range",
"value_type": "integer"
}
],
"type": "rerank"
},
{
"max_batch_size": 100,
"max_sequence_length": 512,
"modality": "text",
"model": "pinecone-rerank-v0",
"provider_name": "Pinecone",
"short_description": "A state of the art reranking model that out-performs competitors on widely accepted benchmarks. It can handle chunks up to 512 tokens (1-2 paragraphs)",
"supported_parameters": [
{
"allowed_values": [
{
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
{
"StringValue": "NONE",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
}
],
"default": {
"StringValue": "END",
"IntValue": null,
"FloatValue": null,
"BoolValue": null
},
"parameter": "truncate",
"required": false,
"type": "one_of",
"value_type": "string"
}
],
"type": "rerank"
}
]
}
```
```csharp C# theme={null}
{
"models": [
{
"model": "llama-text-embed-v2",
"short_description": "A high performance dense embedding model optimized for multilingual and cross-lingual text question-answering retrieval with support for long documents (up to 2048 tokens) and dynamic embedding size (Matryoshka Embeddings).",
"type": "embed",
"vector_type": "dense",
"default_dimension": 1024,
"modality": "text",
"max_sequence_length": 2048,
"max_batch_size": 96,
"provider_name": "NVIDIA",
"supported_dimensions": [
384,
512,
768,
1024,
2048
],
"supported_metrics": [
"cosine",
"cosine"
],
"supported_parameters": [
{
"parameter": "input_type",
"type": "one_of",
"value_type": "string",
"required": true,
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"allowed_values": [
"END",
"NONE",
"START"
],
"default": "END"
},
{
"parameter": "dimension",
"type": "one_of",
"value_type": "integer",
"required": false,
"allowed_values": [
384,
512,
768,
1024,
2048
],
"default": 1024
}
]
},
{
"model": "multilingual-e5-large",
"short_description": "A high-performance dense embedding model trained on a mixture of multilingual datasets. It works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
"type": "embed",
"vector_type": "dense",
"default_dimension": 1024,
"modality": "text",
"max_sequence_length": 507,
"max_batch_size": 96,
"provider_name": "Microsoft",
"supported_dimensions": [
1024
],
"supported_metrics": [
"cosine",
"cosine"
],
"supported_parameters": [
{
"parameter": "input_type",
"type": "one_of",
"value_type": "string",
"required": true,
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"allowed_values": [
"END",
"NONE"
],
"default": "END"
}
]
},
{
"model": "pinecone-sparse-english-v0",
"short_description": "A sparse embedding model for converting text to sparse vectors for keyword or hybrid semantic/keyword search. Built on the innovations of the DeepImpact architecture.",
"type": "embed",
"vector_type": "sparse",
"modality": "text",
"max_sequence_length": 512,
"max_batch_size": 96,
"provider_name": "Pinecone",
"supported_metrics": [
"cosine"
],
"supported_parameters": [
{
"parameter": "input_type",
"type": "one_of",
"value_type": "string",
"required": true,
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"allowed_values": [
"END",
"NONE"
],
"default": "END"
},
{
"parameter": "return_tokens",
"type": "any",
"value_type": "boolean",
"required": false,
"default": false
}
]
},
{
"model": "bge-reranker-v2-m3",
"short_description": "A high-performance, multilingual reranking model that works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
"type": "rerank",
"modality": "text",
"max_sequence_length": 1024,
"max_batch_size": 100,
"provider_name": "BAAI",
"supported_parameters": [
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"allowed_values": [
"END",
"NONE"
],
"default": "NONE"
}
]
},
{
"model": "cohere-rerank-3.5",
"short_description": "Cohere\u0027s leading reranking model, balancing performance and latency for a wide range of enterprise search applications.",
"type": "rerank",
"modality": "text",
"max_sequence_length": 40000,
"max_batch_size": 200,
"provider_name": "Cohere",
"supported_parameters": [
{
"parameter": "max_chunks_per_doc",
"type": "numeric_range",
"value_type": "integer",
"required": false,
"min": 1,
"max": 3072,
"default": 3072
}
]
},
{
"model": "pinecone-rerank-v0",
"short_description": "A state of the art reranking model that out-performs competitors on widely accepted benchmarks. It can handle chunks up to 512 tokens (1-2 paragraphs)",
"type": "rerank",
"modality": "text",
"max_sequence_length": 512,
"max_batch_size": 100,
"provider_name": "Pinecone",
"supported_parameters": [
{
"parameter": "truncate",
"type": "one_of",
"value_type": "string",
"required": false,
"allowed_values": [
"END",
"NONE"
],
"default": "END"
}
]
}
]
}
```
```json curl theme={null}
{
"models": [
{
"model": "llama-text-embed-v2",
"short_description": "A high performance dense embedding model optimized for multilingual and cross-lingual text question-answering retrieval with support for long documents (up to 2048 tokens) and dynamic embedding size (Matryoshka Embeddings).",
"type": "embed",
"vector_type": "dense",
"default_dimension": 1024,
"modality": "text",
"max_sequence_length": 2048,
"max_batch_size": 96,
"provider_name": "NVIDIA",
"supported_metrics": [
"Cosine",
"DotProduct"
],
"supported_dimensions": [
384,
512,
768,
1024,
2048
],
"supported_parameters": [
{
"parameter": "input_type",
"required": true,
"type": "one_of",
"value_type": "string",
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"required": false,
"default": "END",
"type": "one_of",
"value_type": "string",
"allowed_values": [
"END",
"NONE",
"START"
]
},
{
"parameter": "dimension",
"required": false,
"default": 1024,
"type": "one_of",
"value_type": "integer",
"allowed_values": [
384,
512,
768,
1024,
2048
]
}
]
},
{
"model": "multilingual-e5-large",
"short_description": "A high-performance dense embedding model trained on a mixture of multilingual datasets. It works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
"type": "embed",
"vector_type": "dense",
"default_dimension": 1024,
"modality": "text",
"max_sequence_length": 507,
"max_batch_size": 96,
"provider_name": "Microsoft",
"supported_metrics": [
"Cosine",
"Euclidean"
],
"supported_dimensions": [
1024
],
"supported_parameters": [
{
"parameter": "input_type",
"required": true,
"type": "one_of",
"value_type": "string",
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"required": false,
"default": "END",
"type": "one_of",
"value_type": "string",
"allowed_values": [
"END",
"NONE"
]
}
]
},
{
"model": "pinecone-sparse-english-v0",
"short_description": "A sparse embedding model for converting text to sparse vectors for keyword or hybrid semantic/keyword search. Built on the innovations of the DeepImpact architecture.",
"type": "embed",
"vector_type": "sparse",
"modality": "text",
"max_sequence_length": 512,
"max_batch_size": 96,
"provider_name": "Pinecone",
"supported_metrics": [
"DotProduct"
],
"supported_parameters": [
{
"parameter": "input_type",
"required": true,
"type": "one_of",
"value_type": "string",
"allowed_values": [
"query",
"passage"
]
},
{
"parameter": "truncate",
"required": false,
"default": "END",
"type": "one_of",
"value_type": "string",
"allowed_values": [
"END",
"NONE"
]
},
{
"parameter": "return_tokens",
"required": false,
"default": false,
"type": "any",
"value_type": "boolean"
}
]
},
{
"model": "bge-reranker-v2-m3",
"short_description": "A high-performance, multilingual reranking model that works well on messy data and short queries expected to return medium-length passages of text (1-2 paragraphs)",
"type": "rerank",
"modality": "text",
"max_sequence_length": 1024,
"max_batch_size": 100,
"provider_name": "BAAI",
"supported_parameters": [
{
"parameter": "truncate",
"required": false,
"default": "NONE",
"type": "one_of",
"value_type": "string",
"allowed_values": [
"END",
"NONE"
]
}
]
},
{
"model": "cohere-rerank-3.5",
"short_description": "Cohere's leading reranking model, balancing performance and latency for a wide range of enterprise search applications.",
"type": "rerank",
"modality": "text",
"max_sequence_length": 40000,
"max_batch_size": 200,
"provider_name": "Cohere",
"supported_parameters": [
{
"parameter": "max_chunks_per_doc",
"required": false,
"default": 3072,
"type": "numeric_range",
"value_type": "integer",
"min": 1,
"max": 3072
}
]
},
{
"model": "pinecone-rerank-v0",
"short_description": "A state of the art reranking model that out-performs competitors on widely accepted benchmarks. It can handle chunks up to 512 tokens (1-2 paragraphs)",
"type": "rerank",
"modality": "text",
"max_sequence_length": 512,
"max_batch_size": 100,
"provider_name": "Pinecone",
"supported_parameters": [
{
"parameter": "truncate",
"required": false,
"default": "END",
"type": "one_of",
"value_type": "string",
"allowed_values": [
"END",
"NONE"
]
}
]
}
]
}
```
## OpenAPI
````yaml https://raw.githubusercontent.com/pinecone-io/pinecone-api/refs/heads/main/2025-04/inference_2025-04.oas.yaml get /models
openapi: 3.0.3
info:
title: Pinecone Inference API
description: >-
Pinecone is a vector database that makes it easy to search and retrieve
billions of high-dimensional vectors.
contact:
name: Pinecone Support
url: https://support.pinecone.io
email: support@pinecone.io
license:
name: Apache 2.0
url: https://www.apache.org/licenses/LICENSE-2.0
version: 2025-04
servers:
- url: https://api.pinecone.io
description: Production API endpoints
security:
- ApiKeyAuth: []
tags:
- name: Inference
description: Model inference
externalDocs:
description: More Pinecone.io API docs
url: https://docs.pinecone.io/introduction
paths:
/models:
get:
tags:
- Inference
summary: List available models
description: >-
List the embedding and reranking models hosted by Pinecone.
You can use hosted models as an integrated part of Pinecone operations
or for standalone embedding and reranking. For more details, see [Vector
embedding](https://docs.pinecone.io/guides/index-data/indexing-overview#vector-embedding)
and [Rerank
results](https://docs.pinecone.io/guides/search/rerank-results).
operationId: list_models
parameters:
- in: query
name: type
description: Filter models by type ('embed' or 'rerank').
schema:
type: string
example: embed
style: form
- in: query
name: vector_type
description: >-
Filter embedding models by vector type ('dense' or 'sparse'). Only
relevant when `type=embed`.
schema:
type: string
example: sparse
style: form
responses:
'200':
description: The list of available models.
content:
application/json:
schema:
$ref: '#/components/schemas/ModelInfoList'
examples:
multiple-models:
summary: Multiple available models.
value:
models:
- max_batch_size: 96
default_dimension: 256
modality: text
model: example-embedding-model
max_sequence_length: 512
provider_name: Embedding Model Provider
supported_dimensions:
- 256
- 512
short_description: An example embedding model.
supported_metrics:
- cosine
- euclidean
supported_parameters:
- allowed_values:
- value1
- value2
parameter: example_required_param
required: true
type: one_of
value_type: string
- allowed_values:
- value1
- value2
default: value1
parameter: example_param_with_default
required: false
type: one_of
value_type: string
- default: 5
max: 10
min: 0
parameter: example_numeric_range
required: false
type: numeric_range
value_type: integer
type: embed
vector_type: dense
- max_batch_size: 100
modality: text
model: example-reranking-model
max_sequence_length: 1024
provider_name: Reranking Model Provider
short_description: An example reranking model.
supported_parameters:
- default: true
parameter: example_any_value
required: false
type: any
value_type: boolean
type: rerank
'401':
description: 'Unauthorized. Possible causes: Invalid API key.'
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
examples:
unauthorized:
summary: Unauthorized
value:
error:
code: UNAUTHENTICATED
message: Invalid API key.
status: 401
'404':
description: Model not found.
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
examples:
model-not-found:
summary: Model not found
value:
error:
code: NOT_FOUND
message: Model example-model not found.
status: 404
'500':
description: Internal server error.
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponse'
examples:
internal-server-error:
summary: Internal server error
value:
error:
code: UNKNOWN
message: Internal server error
status: 500
components:
schemas:
ModelInfoList:
description: The list of available models.
type: object
properties:
models:
type: array
items:
$ref: '#/components/schemas/ModelInfo'
ErrorResponse:
example:
error:
code: QUOTA_EXCEEDED
message: >-
The index exceeds the project quota of 5 pods by 2 pods. Upgrade
your account or change the project settings to increase the quota.
status: 429
description: The response shape used for all error responses.
type: object
properties:
status:
example: 500
description: The HTTP status code of the error.
type: integer
error:
example:
code: INVALID_ARGUMENT
message: >-
Index name must contain only lowercase alphanumeric characters or
hyphens, and must not begin or end with a hyphen.
description: Detailed information about the error that occurred.
type: object
properties:
code:
type: string
enum:
- OK
- UNKNOWN
- INVALID_ARGUMENT
- DEADLINE_EXCEEDED
- QUOTA_EXCEEDED
- NOT_FOUND
- ALREADY_EXISTS
- PERMISSION_DENIED
- UNAUTHENTICATED
- RESOURCE_EXHAUSTED
- FAILED_PRECONDITION
- ABORTED
- OUT_OF_RANGE
- UNIMPLEMENTED
- INTERNAL
- UNAVAILABLE
- DATA_LOSS
- FORBIDDEN
message:
example: >-
Index name must contain only lowercase alphanumeric characters
or hyphens, and must not begin or end with a hyphen.
type: string
details:
description: >-
Additional information about the error. This field is not
guaranteed to be present.
type: object
required:
- code
- message
required:
- status
- error
ModelInfo:
description: >-
Represents the model configuration including model type, supported
parameters, and other model details.
type: object
properties:
model:
example: multilingual-e5-large
description: The name of the model.
type: string
short_description:
example: multilingual-e5-large
description: A summary of the model.
type: string
type:
example: embed
description: The type of model (e.g. 'embed' or 'rerank').
type: string
vector_type:
description: Whether the embedding model produces 'dense' or 'sparse' embeddings.
type: string
default_dimension:
example: 1024
description: >-
The default embedding model dimension (applies to dense embedding
models only).
type: integer
format: int32
minimum: 1
maximum: 20000
modality:
example: text
description: The modality of the model (e.g. 'text').
type: string
max_sequence_length:
example: 512
description: The maximum tokens per sequence supported by the model.
type: integer
format: int32
minimum: 1
max_batch_size:
example: 96
description: The maximum batch size (number of sequences) supported by the model.
type: integer
format: int32
minimum: 1
provider_name:
example: NVIDIA
description: The name of the provider of the model.
type: string
supported_dimensions:
description: >-
The list of supported dimensions for the model (applies to dense
embedding models only).
type: array
items:
example: 1024
type: integer
format: int32
minimum: 1
maximum: 20000
supported_metrics:
$ref: '#/components/schemas/ModelInfoSupportedMetrics'
supported_parameters:
type: array
items:
$ref: '#/components/schemas/ModelInfoSupportedParameter'
required:
- model
- short_description
- type
- supported_parameters
ModelInfoSupportedMetrics:
description: The distance metrics supported by the model for similarity search.
type: array
items:
$ref: '#/components/schemas/ModelInfoMetric'
ModelInfoSupportedParameter:
description: >-
Describes a parameter supported by the model, including parameter value
constraints.
type: object
properties:
parameter:
example: input_type
description: The name of the parameter.
type: string
type:
example: one_of
description: >-
The parameter type e.g. 'one_of', 'numeric_range', or 'any'.
If the type is 'one_of', then 'allowed_values' will be set, and the
value specified must be one of the allowed values. 'one_of' is only
compatible with value_type 'string' or 'integer'.
If 'numeric_range', then 'min' and 'max' will be set, then the value
specified must adhere to the value_type and must fall within the
`[min, max]` range (inclusive).
If 'any' then any value is allowed, as long as it adheres to the
value_type.
type: string
value_type:
example: string
description: >-
The type of value the parameter accepts, e.g. 'string', 'integer',
'float', or 'boolean'.
type: string
required:
example: true
description: Whether the parameter is required (true) or optional (false).
type: boolean
allowed_values:
description: The allowed parameter values when the type is 'one_of'.
type: array
items:
anyOf:
- type: string
- type: integer
min:
example: 1
description: >-
The minimum allowed value (inclusive) when the type is
'numeric_range'.
type: number
max:
example: 1
description: >-
The maximum allowed value (inclusive) when the type is
'numeric_range'.
type: number
default:
example: END
description: The default value for the parameter when a parameter is optional.
anyOf:
- type: string
- type: integer
format: int32
- type: number
format: float
- type: boolean
required:
- parameter
- type
- value_type
- required
ModelInfoMetric:
description: >-
A distance metric that the embedding model supports for similarity
searches.
type: string
enum:
- cosine
- euclidean
- dotproduct
securitySchemes:
ApiKeyAuth:
type: apiKey
in: header
name: Api-Key
description: >-
An API Key is required to call Pinecone APIs. Get yours from the
[console](https://app.pinecone.io/).
````