`fireworks://`¶

Fireworks AI — fast hosted inference for open-source models (Llama, DeepSeek, Mixtral, FireFunction, Yi, etc.) via an OpenAI-compatible API.

Built on pkg/component/llmutil/ — the same chat/embeddings shape as openai://. This page covers only the Fireworks-specific defaults.

URI shape¶

fireworks://chat?model=<id>&apiKey=<key>&temperature=<float>&maxTokens=<int>&baseURL=<url>
fireworks://embeddings?model=<id>&apiKey=<key>

Parameters¶

Param	Type	Default	Producer/Consumer/Both	Description
path	enum	required	Producer	`chat` or `embeddings`.
`model`	string	`accounts/fireworks/models/llama-v3p1-70b-instruct`	Producer	Fireworks model id — note Fireworks uses fully-qualified `accounts/<org>/models/<name>` ids. Browse the catalogue at https://fireworks.ai/models.
`apiKey`	string	(from constructor)	Producer	Fireworks API key. Sent as `Authorization: Bearer <key>`.
`baseURL`	string	`https://api.fireworks.ai/inference/v1`	Producer	Override for proxies.
`temperature`	float	`0.7`	Producer	Sampling temperature.
`maxTokens`	int	`4096`	Producer	Output cap.

Producer-only. Default timeout 120 s.

Headers¶

Same as openai:// but namespaced under fireworks.* (fireworks.model, fireworks.id, fireworks.finishReason, fireworks.usage.*). Input header systemPrompt works the same way.

Examples¶

- to: fireworks://chat?model=accounts/fireworks/models/llama-v3p1-70b-instruct&apiKey=${FIREWORKS_API_KEY}

DeepSeek for code:

- to: fireworks://chat?model=accounts/fireworks/models/deepseek-v3&apiKey=${FIREWORKS_API_KEY}&temperature=0.1

Behavior notes¶

See openai:// for shared behaviour. The fully-qualified model id (with accounts/... prefix) is the most common gotcha.

Source¶

pkg/component/fireworks/fireworks.go — eight-line wrapper around llmutil.NewComponent.
pkg/component/llmutil/llmutil.go.

fireworks://¶