or run

npx @tessl/cli init
Log in

Version

Tile

Overview

Evals

Files

Files

docs

agents.mdaudio.mdbatch.mdbeta.mdchat-completions.mdclassification.mdembeddings.mdfiles.mdfim.mdfine-tuning.mdindex.mdmodels.mdocr.md

models.mddocs/

0

# Models

1

2

List and manage available models including base models and fine-tuned models. The models API provides access to model metadata, capabilities, and management operations.

3

4

## Capabilities

5

6

### List Models

7

8

Retrieve information about all available models including base models and fine-tuned models.

9

10

```python { .api }

11

def list(**kwargs) -> ModelList:

12

"""

13

List all available models.

14

15

Returns:

16

ModelList containing model information and metadata

17

"""

18

```

19

20

### Retrieve Model

21

22

Get detailed information about a specific model including its capabilities and configuration.

23

24

```python { .api }

25

def retrieve(model_id: str, **kwargs) -> Union[BaseModelCard, FTModelCard]:

26

"""

27

Retrieve detailed information about a specific model.

28

29

Parameters:

30

- model_id: Unique identifier of the model

31

32

Returns:

33

Model card with detailed information (BaseModelCard for base models,

34

FTModelCard for fine-tuned models)

35

"""

36

```

37

38

### Delete Model

39

40

Delete a fine-tuned model that is no longer needed.

41

42

```python { .api }

43

def delete(model_id: str, **kwargs) -> DeleteModelOut:

44

"""

45

Delete a fine-tuned model.

46

47

Parameters:

48

- model_id: Unique identifier of the fine-tuned model to delete

49

50

Returns:

51

Deletion confirmation with model information

52

53

Note: Only fine-tuned models can be deleted, not base models

54

"""

55

```

56

57

## Usage Examples

58

59

### List Available Models

60

61

```python

62

from mistralai import Mistral

63

64

client = Mistral(api_key="your-api-key")

65

66

# Get all available models

67

models = client.models.list()

68

69

print(f"Total models: {len(models.data)}")

70

for model in models.data:

71

print(f"- {model.id}: {model.description or 'No description'}")

72

if hasattr(model, 'capabilities'):

73

print(f" Capabilities: {', '.join(model.capabilities)}")

74

print()

75

```

76

77

### Get Model Details

78

79

```python

80

# Get details for a specific model

81

model_id = "mistral-small-latest"

82

model_info = client.models.retrieve(model_id)

83

84

print(f"Model ID: {model_info.id}")

85

print(f"Type: {model_info.type}")

86

print(f"Created: {model_info.created}")

87

88

if hasattr(model_info, 'max_context_length'):

89

print(f"Max context length: {model_info.max_context_length}")

90

91

if hasattr(model_info, 'capabilities'):

92

print(f"Capabilities: {model_info.capabilities}")

93

94

if hasattr(model_info, 'description'):

95

print(f"Description: {model_info.description}")

96

```

97

98

### Model Filtering and Selection

99

100

```python

101

# Filter models by capabilities

102

models = client.models.list()

103

chat_models = []

104

embedding_models = []

105

106

for model in models.data:

107

if hasattr(model, 'capabilities'):

108

if 'completion' in model.capabilities or 'chat' in model.capabilities:

109

chat_models.append(model)

110

if 'embedding' in model.capabilities:

111

embedding_models.append(model)

112

113

print("Chat/Completion Models:")

114

for model in chat_models:

115

print(f" - {model.id}")

116

117

print("\nEmbedding Models:")

118

for model in embedding_models:

119

print(f" - {model.id}")

120

```

121

122

### Delete Fine-tuned Model

123

124

```python

125

# Delete a fine-tuned model (only works for custom fine-tuned models)

126

try:

127

result = client.models.delete("ft-model-id-example")

128

print(f"Deleted model: {result.id}")

129

print(f"Status: {result.deleted}")

130

except Exception as e:

131

print(f"Error deleting model: {e}")

132

```

133

134

## Types

135

136

### Model List Types

137

138

```python { .api }

139

class ModelList:

140

object: str

141

data: List[Union[BaseModelCard, FTModelCard]]

142

143

class BaseModelCard:

144

id: str

145

object: str

146

created: int

147

owned_by: str

148

type: str

149

description: Optional[str]

150

max_context_length: Optional[int]

151

aliases: Optional[List[str]]

152

capabilities: Optional[List[str]]

153

default_model_temperature: Optional[float]

154

max_temperature: Optional[float]

155

156

class FTModelCard:

157

id: str

158

object: str

159

created: int

160

owned_by: str

161

type: str

162

description: Optional[str]

163

capabilities: Optional[List[str]]

164

job: Optional[str]

165

archived: Optional[bool]

166

```

167

168

### Model Operations

169

170

```python { .api }

171

class DeleteModelOut:

172

id: str

173

object: str

174

deleted: bool

175

```

176

177

### Model Capabilities

178

179

```python { .api }

180

class ModelCapabilities:

181

completion: bool

182

chat: bool

183

embedding: bool

184

fine_tuning: bool

185

function_calling: bool

186

vision: bool

187

```

188

189

## Model Categories

190

191

### Base Models

192

193

**Chat/Completion Models:**

194

- `mistral-large-latest`: Most capable model for complex tasks

195

- `mistral-small-latest`: Efficient model for most tasks

196

- `mistral-medium-latest`: Balanced performance and capability

197

- `codestral-latest`: Specialized for code generation and completion

198

199

**Embedding Models:**

200

- `mistral-embed`: General-purpose text embedding model

201

202

### Fine-tuned Models

203

204

Fine-tuned models have custom identifiers and are created through the fine-tuning API. They inherit capabilities from their base models but are customized for specific use cases.

205

206

### Model Selection Guidelines

207

208

- **mistral-large-latest**: Complex reasoning, analysis, creative tasks

209

- **mistral-small-latest**: General conversations, simple tasks, cost-effective

210

- **codestral-latest**: Code generation, programming assistance, technical documentation

211

- **mistral-embed**: Semantic search, similarity comparison, text classification

212

213

### Model Properties

214

215

- **Context Length**: Maximum number of tokens the model can process in a single request

216

- **Temperature Range**: Supported temperature values for controlling randomness

217

- **Capabilities**: List of supported features (completion, chat, embedding, etc.)

218

- **Aliases**: Alternative names for the same model

219

- **Ownership**: Organization or entity that owns/created the model