> ## Documentation Index
> Fetch the complete documentation index at: https://composio-preview.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Fireworks AI

> Fireworks AI provides APIs for serverless model inference, fine-tuning, deployments, datasets, evaluations, and account management.

![Fireworks AI logo](https://mintcdn.com/composio-preview/GL28i3tBlbEbBd-f/images/toolkits/fireworks_ai.svg?fit=max\&auto=format\&n=GL28i3tBlbEbBd-f\&q=85\&s=f545fcab26e9305b35f2c1bafdd0e563)

* **Category:** artificial intelligence
* **Auth:** API\_KEY
* **Composio-managed OAuth available?** N/A
* **Tools:** 9
* **Triggers:** 0
* **Slug:** `FIREWORKS_AI`
* **Version:** 00000000\_00

## Tools

> [accordion] Create Chat Completion
>
> **Slug:** `FIREWORKS_AI_CREATE_CHAT_COMPLETION`
>
> Generate one non-streaming OpenAI-compatible chat completion from role-based messages using a caller-supplied Fireworks model or deployment ID.
>
> #### Input Parameters
>
> **seed** `integer`
>
> Optional seed for best-effort repeatability.
>
> **stop** `string`
>
> One or more sequences that stop generation.
>
> **model** `string` *required*
>
> Exact Fireworks model or deployed-model identifier to invoke; discover candidates with the Fireworks AI model or deployment listing tools.
>
> **tools** `array`
>
> Function definitions the model may call.
>
> **top\_p** `number`
>
> Nucleus-sampling probability mass; normally tune this or temperature, not both.
>
> **messages** `array` *required*
>
> Conversation messages in order; each item must contain a role and OpenAI-compatible content or tool-call data.
>
> **max\_tokens** `integer`
>
> Maximum number of tokens to generate; omit to use the model or provider default.
>
> **temperature** `number`
>
> Sampling temperature; omit to use the provider default.
>
> **tool\_choice** `string`
>
> Whether and which supplied function tools the model may call.
>
> **response\_format** `object`
>
> Typed output format, such as JSON mode, JSON Schema, grammar, or text, when the model supports it.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Create Embeddings
>
> **Slug:** `FIREWORKS_AI_CREATE_EMBEDDINGS`
>
> Create embedding vectors for one or more caller-provided inputs using a caller-supplied Fireworks embedding model.
>
> #### Input Parameters
>
> **input** `string` *required*
>
> Text, a batch of up to 2048 texts, or arbitrary structured records whose keys are referenced by prompt\_template. For example, input \{'text': 'Hello', 'source': 'docs'} can use prompt\_template 'Embed \{text} from \{source}'.
>
> **model** `string` *required*
>
> Exact embedding model identifier; Fireworks examples may use provider/model form rather than a full account resource name.
>
> **normalize** `boolean`
>
> Whether to normalize returned vectors; omit to use the model default.
>
> **dimensions** `integer`
>
> Requested output vector dimensions when the selected model supports dimension reduction.
>
> **prompt\_template** `string`
>
> Template that substitutes structured-input keys, for example 'Embed this text: \{text}'.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Generate Model Response
>
> **Slug:** `FIREWORKS_AI_GENERATE_MODEL_RESPONSE`
>
> Generate one non-streaming Responses API result, optionally continuing a stored response or using function, MCP, or SSE tools; storage is off by default.
>
> #### Input Parameters
>
> **input** `string` *required*
>
> Text input or ordered input items. Use a plain string for simple prompts; to continue a function-tool loop, replay the model's function\_call item followed by a function\_call\_output item carrying the tool result.
>
> **model** `string` *required*
>
> Exact Fireworks model or deployed-model identifier to invoke; no model is selected by the toolkit.
>
> **store** `boolean`
>
> Whether Fireworks should retain the response for later continuation; defaults to false for ephemeral calls.
>
> **tools** `array`
>
> Verified function, MCP, or SSE tool definitions available to the model.
>
> **top\_p** `number`
>
> Nucleus-sampling probability mass from 0 to 1; normally tune this or temperature, not both.
>
> **truncation** `string ("auto" | "disabled")`
>
> Use 'auto' to truncate overlong context or 'disabled' to return an error; omit for the provider default.
>
> **temperature** `number`
>
> Sampling temperature from 0 to 2; omit to use the provider default.
>
> **tool\_choice** `string`
>
> Whether the model must avoid tools, choose automatically, call at least one supplied tool, or call one specific named function tool.
>
> **instructions** `string`
>
> System-level instructions that guide the model for this response.
>
> **max\_tool\_calls** `integer`
>
> Maximum number of tool calls allowed for this response.
>
> **max\_output\_tokens** `integer`
>
> Maximum number of output tokens across generated content; omit to use the model limit.
>
> **previous\_response\_id** `string`
>
> ID of a previously stored response to continue; omit for an independent response.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Deployment
>
> **Slug:** `FIREWORKS_AI_GET_DEPLOYMENT`
>
> Return configuration and serving status for one Fireworks deployment without scaling or modifying it.
>
> #### Input Parameters
>
> **read\_mask** `string`
>
> Comma-separated deployment fields to return; omit for the provider's default projection.
>
> **account\_id** `string` *required*
>
> Fireworks account ID containing the deployment.
>
> **deployment\_id** `string` *required*
>
> Deployment resource ID within the account, obtainable with the Fireworks AI deployment listing tool.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Model
>
> **Slug:** `FIREWORKS_AI_GET_MODEL`
>
> Return metadata and status for one model resource in a Fireworks account; this does not test whether the model can serve inference.
>
> #### Input Parameters
>
> **model\_id** `string` *required*
>
> Model resource ID within the account, obtainable with the Fireworks AI model listing tool.
>
> **read\_mask** `string`
>
> Comma-separated model fields to return; omit for the provider's default projection.
>
> **account\_id** `string` *required*
>
> Fireworks account ID containing the model resource.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] List Accounts
>
> **Slug:** `FIREWORKS_AI_LIST_ACCOUNTS`
>
> Return one page of Fireworks accounts accessible to the API key so an agent can obtain account IDs for model and deployment discovery.
>
> #### Input Parameters
>
> **filter** `string`
>
> Fireworks management API filter expression for accounts.
>
> **page\_size** `integer`
>
> Maximum accounts to return in this page (1-200).
>
> **read\_mask** `string`
>
> Comma-separated account fields to return; omit for the provider's default projection.
>
> **next\_cursor** `string`
>
> Continuation cursor returned by a previous call to this tool; omit for the first page.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] List Deployments
>
> **Slug:** `FIREWORKS_AI_LIST_DEPLOYMENTS`
>
> Return one page of deployment resources in a Fireworks account for deployment ID discovery and serving-capacity and status inspection.
>
> #### Input Parameters
>
> **filter** `string`
>
> Fireworks management API filter expression for deployment resources.
>
> **order\_by** `string`
>
> Comma-separated deployment fields to order by; append ' desc' for descending order.
>
> **page\_size** `integer`
>
> Maximum deployment resources to return in this page (1-200).
>
> **read\_mask** `string`
>
> Comma-separated deployment fields to return; omit for the provider's default projection.
>
> **account\_id** `string` *required*
>
> Fireworks account ID, obtainable with the Fireworks AI account listing tool.
>
> **next\_cursor** `string`
>
> Continuation cursor returned by a previous call to this tool; omit for the first page.
>
> **show\_deleted** `boolean`
>
> Whether to include soft-deleted deployment resources in discovery results.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] List Models
>
> **Slug:** `FIREWORKS_AI_LIST_MODELS`
>
> Return one page of model resources in a Fireworks account for model-ID discovery and status inspection; availability for inference is not guaranteed.
>
> #### Input Parameters
>
> **filter** `string`
>
> Fireworks management API filter expression for model resources.
>
> **order\_by** `string`
>
> Comma-separated model fields to order by; append ' desc' for descending order.
>
> **page\_size** `integer`
>
> Maximum model resources to return in this page (1-200).
>
> **read\_mask** `string`
>
> Comma-separated model fields to return; omit for the provider's default projection.
>
> **account\_id** `string` *required*
>
> Fireworks account ID, obtainable with the Fireworks AI account listing tool.
>
> **next\_cursor** `string`
>
> Continuation cursor returned by a previous call to this tool; omit for the first page.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Rerank Documents
>
> **Slug:** `FIREWORKS_AI_RERANK_DOCUMENTS`
>
> Rank a provided document list by relevance to a query using an optional caller-supplied Fireworks reranker model.
>
> #### Input Parameters
>
> **task** `string`
>
> Optional task instruction supported by the selected reranker model.
>
> **model** `string`
>
> Exact reranker model identifier; omit only when intentionally using Fireworks' documented provider default.
>
> **query** `string` *required*
>
> Search query against which document relevance is scored.
>
> **top\_n** `integer`
>
> Maximum number of highest-ranked results to return.
>
> **documents** `array` *required*
>
> Documents to score and reorder; result indices refer to this original array.
>
> **return\_documents** `boolean`
>
> Whether each result should include its document text as well as its original index and relevance score.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not
