vw_model_capabilities
Creates, updates, deletes, gets or lists a vw_model_capabilities resource.
Overview
| Name | vw_model_capabilities |
| Type | View |
| Id | gemini.models.vw_model_capabilities |
Fields
The following fields are returned by this view:
| Name | Datatype | Description |
|---|---|---|
name | string | Model resource name, e.g. models/gemini-2.5-flash. |
display_name | string | Human-readable name of the model. |
version | string | Model version string. |
input_token_limit | integer | Maximum number of input tokens allowed. |
output_token_limit | integer | Maximum number of output tokens available. |
supports_generate_content | boolean | Whether the model supports generateContent. |
supports_count_tokens | boolean | Whether the model supports countTokens. |
supports_embed_content | boolean | Whether the model supports embedContent. |
supports_batching | boolean | Whether the model supports batchGenerateContent. |
supports_caching | boolean | Whether the model supports createCachedContent. |
thinking | boolean | Whether the model supports extended thinking. |
temperature | number | Default sampling temperature. |
max_temperature | number | Maximum sampling temperature the model accepts. |
top_p | number | Default nucleus sampling probability mass. |
top_k | integer | Default top-k sampling cutoff. |
SELECT Examples
SELECT
name,
display_name,
version,
input_token_limit,
output_token_limit,
supports_generate_content,
supports_count_tokens,
supports_embed_content,
supports_batching,
supports_caching,
thinking,
temperature,
max_temperature,
top_p,
top_k
FROM gemini.models.vw_model_capabilities;
SQL Definition
- Sqlite3
- Postgres
SELECT
name,
displayName AS display_name,
version,
inputTokenLimit AS input_token_limit,
outputTokenLimit AS output_token_limit,
CASE WHEN supportedGenerationMethods LIKE '%generateContent%' THEN 1 ELSE 0 END AS supports_generate_content,
CASE WHEN supportedGenerationMethods LIKE '%countTokens%' THEN 1 ELSE 0 END AS supports_count_tokens,
CASE WHEN supportedGenerationMethods LIKE '%embedContent%' THEN 1 ELSE 0 END AS supports_embed_content,
CASE WHEN supportedGenerationMethods LIKE '%batchGenerateContent%' THEN 1 ELSE 0 END AS supports_batching,
CASE WHEN supportedGenerationMethods LIKE '%createCachedContent%' THEN 1 ELSE 0 END AS supports_caching,
thinking,
temperature,
maxTemperature AS max_temperature,
topP AS top_p,
topK AS top_k
FROM gemini.models.models
SELECT
name,
displayName AS display_name,
version,
inputTokenLimit AS input_token_limit,
outputTokenLimit AS output_token_limit,
CASE WHEN supportedGenerationMethods LIKE '%generateContent%' THEN 1 ELSE 0 END AS supports_generate_content,
CASE WHEN supportedGenerationMethods LIKE '%countTokens%' THEN 1 ELSE 0 END AS supports_count_tokens,
CASE WHEN supportedGenerationMethods LIKE '%embedContent%' THEN 1 ELSE 0 END AS supports_embed_content,
CASE WHEN supportedGenerationMethods LIKE '%batchGenerateContent%' THEN 1 ELSE 0 END AS supports_batching,
CASE WHEN supportedGenerationMethods LIKE '%createCachedContent%' THEN 1 ELSE 0 END AS supports_caching,
thinking,
temperature,
maxTemperature AS max_temperature,
topP AS top_p,
topK AS top_k
FROM gemini.models.models