Skip to content
Source
Sign inSign up
TanStack/ai
Open History to see the last commit

TextOptions.md

226 lines5.5 KB

1---
2id: TextOptions
3title: TextOptions
4---
5​
6# Interface: TextOptions\<TProviderOptionsSuperset, TProviderOptionsForModel\>
7​
8Defined in: [types.ts:545](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L545)
9​
10Options passed into the SDK and further piped to the AI provider.
11​
12## Type Parameters
13​
14### TProviderOptionsSuperset
15​
16`TProviderOptionsSuperset` *extends* `Record`\<`string`, `any`\> = `Record`\<`string`, `any`\>
17​
18### TProviderOptionsForModel
19​
20`TProviderOptionsForModel` = `TProviderOptionsSuperset`
21​
22## Properties
23​
24### abortController?
25​
26```ts
27optional abortController: AbortController;
28```
29​
30Defined in: [types.ts:628](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L628)
31​
32AbortController for request cancellation.
33​
34Allows you to cancel an in-progress request using an AbortController.
35Useful for implementing timeouts or user-initiated cancellations.
36​
37#### Example
38​
39```ts
40const abortController = new AbortController();
41setTimeout(() => abortController.abort(), 5000); // Cancel after 5 seconds
42await chat({ ..., abortController });
43```
44​
45#### See
46​
47https://developer.mozilla.org/en-US/docs/Web/API/AbortController
48​
49***
50​
51### agentLoopStrategy?
52​
53```ts
54optional agentLoopStrategy: AgentLoopStrategy;
55```
56​
57Defined in: [types.ts:553](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L553)
58​
59***
60​
61### conversationId?
62​
63```ts
64optional conversationId: string;
65```
66​
67Defined in: [types.ts:614](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L614)
68​
69Conversation ID for correlating client and server-side devtools events.
70When provided, server-side events will be linked to the client conversation in devtools.
71​
72***
73​
74### maxTokens?
75​
76```ts
77optional maxTokens: number;
78```
79​
80Defined in: [types.ts:588](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L588)
81​
82The maximum number of tokens to generate in the response.
83​
84Provider usage:
85- OpenAI: `max_output_tokens` (number) - includes visible output and reasoning tokens
86- Anthropic: `max_tokens` (number, required) - range x >= 1
87- Gemini: `generationConfig.maxOutputTokens` (number)
88​
89***
90​
91### messages
92​
93```ts
94messages: ModelMessage<
95 | string
96 | ContentPart<unknown, unknown, unknown, unknown, unknown>[]
97 | null>[];
98```
99​
100Defined in: [types.ts:550](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L550)
101​
102***
103​
104### metadata?
105​
106```ts
107optional metadata: Record<string, any>;
108```
109​
110Defined in: [types.ts:599](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L599)
111​
112Additional metadata to attach to the request.
113Can be used for tracking, debugging, or passing custom information.
114Structure and constraints vary by provider.
115​
116Provider usage:
117- OpenAI: `metadata` (Record<string, string>) - max 16 key-value pairs, keys max 64 chars, values max 512 chars
118- Anthropic: `metadata` (Record<string, any>) - includes optional user_id (max 256 chars)
119- Gemini: Not directly available in TextProviderOptions
120​
121***
122​
123### model
124​
125```ts
126model: string;
127```
128​
129Defined in: [types.ts:549](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L549)
130​
131***
132​
133### modelOptions?
134​
135```ts
136optional modelOptions: TProviderOptionsForModel;
137```
138​
139Defined in: [types.ts:600](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L600)
140​
141***
142​
143### outputSchema?
144​
145```ts
146optional outputSchema: ZodType<unknown, unknown, $ZodTypeInternals<unknown, unknown>>;
147```
148​
149Defined in: [types.ts:609](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L609)
150​
151Zod schema for structured output.
152When provided, the adapter should use the provider's native structured output API
153to ensure the response conforms to this schema.
154The schema will be converted to JSON Schema format before being sent to the provider.
155​
156***
157​
158### request?
159​
160```ts
161optional request: Request | RequestInit;
162```
163​
164Defined in: [types.ts:601](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L601)
165​
166***
167​
168### systemPrompts?
169​
170```ts
171optional systemPrompts: string[];
172```
173​
174Defined in: [types.ts:552](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L552)
175​
176***
177​
178### temperature?
179​
180```ts
181optional temperature: number;
182```
183​
184Defined in: [types.ts:566](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L566)
185​
186Controls the randomness of the output.
187Higher values (e.g., 0.8) make output more random, lower values (e.g., 0.2) make it more focused and deterministic.
188Range: [0.0, 2.0]
189​
190Note: Generally recommended to use either temperature or topP, but not both.
191​
192Provider usage:
193- OpenAI: `temperature` (number) - in text.top_p field
194- Anthropic: `temperature` (number) - ranges from 0.0 to 1.0, default 1.0
195- Gemini: `generationConfig.temperature` (number) - ranges from 0.0 to 2.0
196​
197***
198​
199### tools?
200​
201```ts
202optional tools: Tool<any, any, any>[];
203```
204​
205Defined in: [types.ts:551](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L551)
206​
207***
208​
209### topP?
210​
211```ts
212optional topP: number;
213```
214​
215Defined in: [types.ts:579](https://github.com/TanStack/ai/blob/main/packages/typescript/ai/src/types.ts#L579)
216​
217Nucleus sampling parameter. An alternative to temperature sampling.
218The model considers the results of tokens with topP probability mass.
219For example, 0.1 means only tokens comprising the top 10% probability mass are considered.
220​
221Note: Generally recommended to use either temperature or topP, but not both.
222​
223Provider usage:
224- OpenAI: `text.top_p` (number)
225- Anthropic: `top_p` (number | null)
226- Gemini: `generationConfig.topP` (number)