From 9838fb02189829b8a63bb7568e7c8e6557f55b8d Mon Sep 17 00:00:00 2001 From: Ivan Despot <66276597+g-despot@users.noreply.github.com> Date: Fri, 14 Aug 2026 13:51:11 +0200 Subject: [PATCH] docs(model-providers): replace hand-maintained Cohere model lists and fix related drift Core stopped validating model names for the Cohere modules in v1.33.0 (backported to v1.31.17 and v1.32.10), so any model string is forwarded verbatim and the static lists on those pages are guaranteed to rot. Replace each list with the no-validation statement, the server default, and a link to Cohere's own model documentation. Also corrected while in these pages: - openai/embeddings.md said you can use any OpenAI embedding model, but text2vec-openai still validates against an allowlist. The page also never stated its server default (text-embedding-3-small), and its "(default)" markers refer to dimensions, not model choice. - openai/generative.md's list is Weaviate's token-limit table, not an availability list. Kept and demoted, with a lead-in saying what it does: it caps maxTokens for those models. - return_likelihoods dropped from four snippets. The Python client accepts and silently discards it, and core has no reader on any path. - Six Python/TypeScript snippet pairs showed different values for the same example (dimensions, the multimodal model, and one model left active in TypeScript while commented in Python). Co-Authored-By: Claude Opus 5 (1M context) Claude-Session: https://claude.ai/code/session_017AryMRLot3XAaz2s44eNG6 --- .../_includes/provider.generative.py | 2 -- .../_includes/provider.generative.ts | 5 +--- .../_includes/provider.vectorizer.py | 5 ++-- .../_includes/provider.vectorizer.ts | 4 +-- .../cohere/embeddings-multimodal.md | 14 ++++----- .../model-providers/cohere/embeddings.md | 30 +++++-------------- .../model-providers/cohere/generative.md | 19 ++---------- .../model-providers/cohere/reranker.md | 14 ++++----- .../model-providers/openai/embeddings.md | 12 ++++---- .../model-providers/openai/generative.md | 14 +++++---- 10 files changed, 43 insertions(+), 76 deletions(-) diff --git a/docs/weaviate/model-providers/_includes/provider.generative.py b/docs/weaviate/model-providers/_includes/provider.generative.py index 4cd4d43f..62bcd540 100644 --- a/docs/weaviate/model-providers/_includes/provider.generative.py +++ b/docs/weaviate/model-providers/_includes/provider.generative.py @@ -383,7 +383,6 @@ def import_data(): # max_tokens=500, # k=5, # stop_sequences=["\n\n"], - # return_likelihoods="GENERATION" ) # highlight-end # Additional parameters not shown @@ -411,7 +410,6 @@ def import_data(): # max_tokens=500, # k=5, # stop_sequences=["\n\n"], - # return_likelihoods="GENERATION" ), # Additional parameters not shown # highlight-end diff --git a/docs/weaviate/model-providers/_includes/provider.generative.ts b/docs/weaviate/model-providers/_includes/provider.generative.ts index e3ddc4ae..86ebe516 100644 --- a/docs/weaviate/model-providers/_includes/provider.generative.ts +++ b/docs/weaviate/model-providers/_includes/provider.generative.ts @@ -343,12 +343,11 @@ await client.collections.create({ // highlight-start generative: weaviate.configure.generative.cohere({ // These parameters are optional - model: 'command-r', + // model: 'command-r', // temperature: 0.7, // maxTokens: 500, // k: 5, // stopSequences: ['\n\n'], - // returnLikelihoods: 'GENERATION' }), // highlight-end // Additional parameters not shown @@ -366,8 +365,6 @@ response = await myCollection.generate.nearText("A holiday film", { // maxTokens: 500, // k: 5, // stopSequences: ['\n\n'], - // returnLikelihoods: 'GENERATION' // coming soon - }), // highlight-end }, { diff --git a/docs/weaviate/model-providers/_includes/provider.vectorizer.py b/docs/weaviate/model-providers/_includes/provider.vectorizer.py index 1acbf121..a2e784c4 100644 --- a/docs/weaviate/model-providers/_includes/provider.vectorizer.py +++ b/docs/weaviate/model-providers/_includes/provider.vectorizer.py @@ -136,7 +136,7 @@ source_properties=["title"], # Further options # model="embed-v4.0", - # dimensions=512, + # dimensions=1024, # truncate="END", # "NONE", "START" or "END" # base_url="" ) @@ -193,7 +193,6 @@ Configure.Vectors.multi2vec_cohere( name="title_vector", model="embed-v4.0", - dimensions=1024, # Define the fields to be used for the vectorization - using image_fields, text_fields image_fields=[ Multi2VecField(name="poster", weight=0.9) @@ -1042,7 +1041,7 @@ # model="text-embedding-3-large", # model_version="002", # Parameter only applicable for `ada` model family and older # dimensions=1024, # Parameter only applicable for `v3` model family and newer - # type="text", + # type_="text", # base_url="", ) ], diff --git a/docs/weaviate/model-providers/_includes/provider.vectorizer.ts b/docs/weaviate/model-providers/_includes/provider.vectorizer.ts index 11f5bef7..13015f95 100644 --- a/docs/weaviate/model-providers/_includes/provider.vectorizer.ts +++ b/docs/weaviate/model-providers/_includes/provider.vectorizer.ts @@ -229,7 +229,7 @@ await client.collections.create({ vectorizers: [ weaviate.configure.vectors.multi2VecCohere({ name: "title_vector", - model: "embed-multilingual-v3.0", + model: "embed-v4.0", // Define the fields to be used for the vectorization - using imageFields, textFields imageFields: [{ name: "poster", @@ -275,7 +275,7 @@ await client.collections.create({ weight: 0.1 }], // Further options - // model: "embed-multilingual-v3.0", + // model: "embed-v4.0", // dimensions: 1024, // truncate: "END", // "NONE", "START" or "END" // baseURL: "" diff --git a/docs/weaviate/model-providers/cohere/embeddings-multimodal.md b/docs/weaviate/model-providers/cohere/embeddings-multimodal.md index e7126e2d..30b91c0f 100644 --- a/docs/weaviate/model-providers/cohere/embeddings-multimodal.md +++ b/docs/weaviate/model-providers/cohere/embeddings-multimodal.md @@ -103,7 +103,7 @@ Provide the API key to Weaviate using one of the following methods: ### Select a model -You can specify one of the [available models](#available-models) for the vectorizer to use, as shown in the following configuration example. +You can specify which [model](#available-models) the vectorizer uses, as shown in the following configuration example. @@ -126,7 +126,7 @@ You can specify one of the [available models](#available-models) for the vectori -You can [specify](#vectorizer-parameters) one of the [available models](#available-models) for Weaviate to use. The [default model](#available-models) is used if no model is specified. +You can [specify](#vectorizer-parameters) which [model](#available-models) Weaviate uses. If you do not specify a model, Weaviate uses the server default. import VectorizationBehavior from '/_includes/vectorization.behavior.mdx'; @@ -294,11 +294,11 @@ The query below returns the `n` most similar objects to the input image from the ### Available models -- `embed-v4.0` -- `embed-multilingual-v3.0` (server default) -- `embed-multilingual-light-v3.0` -- `embed-english-v3.0` -- `embed-english-light-v3.0` +Weaviate does not validate the model name, so you can set any model that your Cohere account can reach. Name validation was removed in `v1.33.0`, and backported to `v1.31.17` and `v1.32.10`. + +The server default is `embed-multilingual-v3.0`. + +A multimodal collection requires a model that can embed images as well as text. See the [Cohere model documentation](https://docs.cohere.com/docs/models) for the list of available models. ## Further resources diff --git a/docs/weaviate/model-providers/cohere/embeddings.md b/docs/weaviate/model-providers/cohere/embeddings.md index 6464cb67..a06d9bd3 100644 --- a/docs/weaviate/model-providers/cohere/embeddings.md +++ b/docs/weaviate/model-providers/cohere/embeddings.md @@ -134,7 +134,7 @@ Provide the API key to Weaviate using one of the following methods: ### Select a model -You can specify one of the [available models](#available-models) for the vectorizer to use, as shown in the following configuration example. +You can specify which [model](#available-models) the vectorizer uses, as shown in the following configuration example. @@ -167,7 +167,7 @@ You can specify one of the [available models](#available-models) for the vectori -You can [specify](#vectorizer-parameters) one of the [available models](#available-models) for Weaviate to use. The [default model](#available-models) is used if no model is specified. +You can [specify](#vectorizer-parameters) which [model](#available-models) Weaviate uses. If you do not specify a model, Weaviate uses the server default. import VectorizationBehavior from '/_includes/vectorization.behavior.mdx'; @@ -355,33 +355,19 @@ The query below returns the `n` best scoring objects from the database, set by ` ### Available models -- `embed-v4.0` -- `embed-multilingual-v3.0` (server default) -- `embed-multilingual-light-v3.0` -- `embed-multilingual-v2.0` (previously `embed-multilingual-22-12`) -- `embed-english-v3.0` -- `embed-english-light-v3.0` -- `embed-english-v2.0` -- `embed-english-light-v2.0` +Weaviate does not validate the model name, so you can set any model that your Cohere account can reach. Name validation was removed in `v1.33.0`, and backported to `v1.31.17` and `v1.32.10`. -
- Deprecated models - -The following models are available, but deprecated: -- `multilingual-22-12` -- `large` -- `medium` -- `small` +The server default is `embed-multilingual-v3.0`. -
+See the [Cohere model documentation](https://docs.cohere.com/docs/models) for the list of available models. ## Further resources ### Other integrations -- [Cohere multimodal embedding models + Weaviate](./embeddings-multimodal.md) -- [Cohere generative models + Weaviate](./generative.md) -- [Cohere reranker models + Weaviate](./reranker.md) +- [Cohere multimodal embedding models + Weaviate](./embeddings-multimodal.md). +- [Cohere generative models + Weaviate](./generative.md). +- [Cohere reranker models + Weaviate](./reranker.md). ### Code examples diff --git a/docs/weaviate/model-providers/cohere/generative.md b/docs/weaviate/model-providers/cohere/generative.md index 15ba4486..775b0498 100644 --- a/docs/weaviate/model-providers/cohere/generative.md +++ b/docs/weaviate/model-providers/cohere/generative.md @@ -108,7 +108,7 @@ import MutableGenerativeConfig from '/_includes/mutable-generative-config.md'; ### Select a model -You can specify one of the [available models](#available-models) for Weaviate to use, as shown in the following configuration example: +You can specify which [model](#available-models) Weaviate uses, as shown in the following configuration example: @@ -131,7 +131,7 @@ You can specify one of the [available models](#available-models) for Weaviate to -You can [specify](#generative-parameters) one of the [available models](#available-models) for Weaviate to use. The [default model](#available-models) is used if no model is specified. +You can [specify](#generative-parameters) which [model](#available-models) Weaviate uses. If you do not specify a model, Weaviate uses the server default. ### Generative parameters @@ -268,20 +268,7 @@ Weaviate does not validate the model name, so you can set any model that your Co The server default is `command-a-03-2025`. It changed in `v1.33.0`, and was backported to `v1.31.17` and `v1.32.10`. Earlier releases on each of those lines default to `command-r`. -The following models are commonly used: - -- `command-a-03-2025` (server default) -- `command-r-plus` -- `command-r` (previous server default) -- `command-xlarge` -- `command-xlarge-beta` -- `command-xlarge-nightly` -- `command-medium` -- `command-medium-nightly` -- `command` -- `command-nightly` -- `command-light` -- `command-light-nightly` +See the [Cohere model documentation](https://docs.cohere.com/docs/models) for the list of available models. ## Further resources diff --git a/docs/weaviate/model-providers/cohere/reranker.md b/docs/weaviate/model-providers/cohere/reranker.md index 5f3565b5..75f1b117 100644 --- a/docs/weaviate/model-providers/cohere/reranker.md +++ b/docs/weaviate/model-providers/cohere/reranker.md @@ -108,7 +108,7 @@ Configure a Weaviate collection to use a Cohere reranker model as follows: ### Select a model -You can specify one of the [available models](#available-models) for Weaviate to use, as shown in the following configuration example: +You can specify which [model](#available-models) Weaviate uses, as shown in the following configuration example: @@ -180,17 +180,13 @@ Any search in Weaviate can be combined with a reranker to perform reranking oper ### Available models -- `rerank-v3.5` (server default) -- `rerank-english-v3.0` -- `rerank-multilingual-v3.0` -- `rerank-english-v2.0` -- `rerank-multilingual-v2.0` +Weaviate does not validate the model name, so you can set any model that your Cohere account can reach. Name validation was removed in `v1.33.0`, and backported to `v1.31.17` and `v1.32.10`. -You can also select a fine-tuned reranker model_id, such as: +The server default is `rerank-v3.5`. -- `500df123-afr3-...` +See the [Cohere model documentation](https://docs.cohere.com/docs/models) for the list of available models. -See [this blog post](https://weaviate.io/blog/fine-tuning-coheres-reranker) for more information. +You can also specify a fine-tuned reranker by its model ID, for example `500df123-afr3-...`. For details, see [Fine-Tuning Cohere's Reranker](https://weaviate.io/blog/fine-tuning-coheres-reranker). For further details on model parameters, see the [Cohere API documentation](https://docs.cohere.com/reference/rerank). diff --git a/docs/weaviate/model-providers/openai/embeddings.md b/docs/weaviate/model-providers/openai/embeddings.md index 51f6cc47..11e68058 100644 --- a/docs/weaviate/model-providers/openai/embeddings.md +++ b/docs/weaviate/model-providers/openai/embeddings.md @@ -129,7 +129,7 @@ Provide the API key to Weaviate using one of the following methods: ### Select a model -You can specify one of the [available models](#available-models) for the vectorizer to use, as shown in the following configuration examples. +You can specify one of the [available models](#available-models) for the vectorizer to use, as shown in the following configuration examples. If you do not set a model, Weaviate uses the server default, `text-embedding-3-small`. #### For `text-embedding-3` model family @@ -199,7 +199,7 @@ For older models such as `text-embedding-ada-002`, provide the model name (`ada` -You can [specify](#vectorizer-parameters) one of the [available models](#available-models) for Weaviate to use. The [default model](#available-models) is used if no model is specified. +You can [specify](#vectorizer-parameters) one of the [available models](#available-models) for Weaviate to use. If no model is specified, Weaviate uses `text-embedding-3-small`. import VectorizationBehavior from '/_includes/vectorization.behavior.mdx'; @@ -212,7 +212,7 @@ import VectorizationBehavior from '/_includes/vectorization.behavior.mdx'; ### Vectorizer parameters -- `model`: The OpenAI model name or family. +- `model`: The OpenAI model name or family. Defaults to `text-embedding-3-small`. - `dimensions`: The number of dimensions for the model. - `modelVersion`: The version string for the model. - `type`: The model type, either `text` or `code`. @@ -412,10 +412,12 @@ The query below returns the `n` best scoring objects from the database, set by ` ### Available models -You can use any OpenAI embedding model with `text2vec-openai`. For document embeddings, choose from the following [embedding model families](https://platform.openai.com/docs/models/embeddings): +The server default is `text-embedding-3-small`. + +For document embeddings, choose from the following [embedding model families](https://platform.openai.com/docs/models/embeddings): * `text-embedding-3` - * Available dimensions: + * Available `dimensions` values: * `text-embedding-3-large`: `256`, `1024`, `3072` (default) * `text-embedding-3-small`: `512`, `1536` (default) * `ada` diff --git a/docs/weaviate/model-providers/openai/generative.md b/docs/weaviate/model-providers/openai/generative.md index b305241c..1bc4446f 100644 --- a/docs/weaviate/model-providers/openai/generative.md +++ b/docs/weaviate/model-providers/openai/generative.md @@ -285,7 +285,7 @@ In other words, when you have `n` search results, the generative model generates ### RAG with images -You can also supply images as a part of the input when performing retrieval augmented generation in both single prompts and grouped tasks. +You can also supply images as a part of the input when performing retrieval augmented generation in both single prompts and grouped tasks. @@ -315,7 +315,12 @@ Weaviate does not validate the model name, so you can set any model that your Op The server default is `gpt-5-mini`. It changed in `v1.32.3`, and was backported to `v1.30.16` and `v1.31.10`. Earlier releases on each of those lines default to `gpt-3.5-turbo`. -The following models are recognized by Weaviate's token limit table: +See the [OpenAI model documentation](https://platform.openai.com/docs/models) for the list of available models. + +Weaviate stores a token limit for the models below. The limit caps the `maxTokens` value you can set for those models; it does not restrict which models you can use. + +
+ Models with a stored token limit * [gpt-5](https://platform.openai.com/docs/models/gpt-5) * [gpt-5-mini](https://platform.openai.com/docs/models/gpt-5-mini) (server default) @@ -329,10 +334,7 @@ The following models are recognized by Weaviate's token limit table: * [gpt-4o](https://platform.openai.com/docs/models#gpt-4o) * [gpt-4o-mini](https://platform.openai.com/docs/models#gpt-4o-mini) -
- Older models - -The following models are available, but not recommended: +These older models also have a stored limit, but are not recommended: * [davinci 002](https://platform.openai.com/docs/models/overview) * [davinci 003](https://platform.openai.com/docs/models/overview)