groq/options

9 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

Groq reasoningFormat option values

The `reasoningFormat` option for Groq reasoning models accepts three values: `'parsed'`, `'raw'`, and `'hidden'`. It controls how reasoning is exposed in generated text and is only supported by reasoning models like `qwen-qwq-32b` and `deepseek-r1-distill-*` models.

Groq reasoningEffort option configuration

The `reasoningEffort` option controls the level of reasoning effort. For `qwen/qwen3-32b`, supported values are `'none'` (disable reasoning) and `'default'` (enable reasoning). For `gpt-oss20b/gpt-oss120b`, supported values are `'low'`, `'medium'`, and `'high'`. Defaults to `'default'` for `qwen/qwen3-32b`.

Groq structuredOutputs option

The `structuredOutputs` option (boolean) determines whether to use structured outputs. Defaults to `true`. When enabled, object generation uses `json_schema` format instead of `json_object` format for more reliable structured outputs.

Groq strictJsonSchema option

The `strictJsonSchema` option (boolean) enables strict JSON schema validation using constrained decoding to guarantee schema compliance. Defaults to `true`. Only used when `structuredOutputs` is enabled and a schema is provided.

Groq parallelToolCalls option

The `parallelToolCalls` option (boolean) enables parallel function calling during tool use. Defaults to `true`.

Groq user option

The `user` option (string) accepts a unique identifier representing an end-user, which can help with monitoring and abuse detection.

Groq serviceTier option values

The `serviceTier` option accepts four values: `'on_demand'` (default tier with consistent performance), `'performance'` (prioritized tier for latency-sensitive workloads), `'flex'` (higher throughput tier with 10x rate limits, handles occasional failures), and `'auto'` (uses on_demand first, falls back to flex if exceeded). Defaults to `'on_demand'`.

Groq language model example with reasoning and options

Example using Groq reasoning model with provider options: `const result = await generateText({ model: groq('qwen/qwen3-32b'), providerOptions: { groq: { reasoningFormat: 'parsed', reasoningEffort: 'default', parallelToolCalls: true, user: 'user-123', serviceTier: 'flex' } satisfies GroqLanguageModelChatOptions }, prompt: 'How many "r"s are in the word "strawberry"?' });`

Groq structured outputs disabled example

Example disabling structured outputs for models that don't support them: `const result = await generateText({ model: groq('gemma2-9b-it'), providerOptions: { groq: { structuredOutputs: false } satisfies GroqLanguageModelChatOptions }, output: Output.object({ schema: z.object({ recipe: z.object({ name: z.string(), ingredients: z.array(z.string()), instructions: z.array(z.string()) }) }) }), prompt: 'Generate a simple pasta recipe in JSON format.' });`

Give your agent this brain