Create a chat completion | OpenRouter

Sends a request for a model response for the given chat conversation. Supports both streaming and non-streaming modes.

Authentication

AuthorizationBearer

API key as bearer token in Authorization header

Request

Chat completion request parameters

messageslist of objectsRequired

providerobject or nullOptional

When multiple model providers are available, optionally indicate your routing preference.

pluginslist of objectsOptional

Plugins you want to enable for this request, including their settings.

routeenum or nullOptional

Routing strategy for multiple models: “fallback” (default) uses secondary models as backups, “sort” sorts all endpoints together by routing criteria.

Allowed values:

userstringOptional

session_idstringOptional<=128 characters

A unique identifier for grouping related requests (e.g., a conversation or agent workflow) for observability. If provided in both the request body and the x-session-id header, the body value takes precedence. Maximum of 128 characters.

modelstringOptional

modelslist of stringsOptional

frequency_penaltydouble or nullOptional-2-2

logit_biasmap from strings to doubles or nullOptional

logprobsboolean or nullOptional

top_logprobsdouble or nullOptional0-20

max_completion_tokensdouble or nullOptional>=1

max_tokensdouble or nullOptional>=1

metadatamap from strings to stringsOptional

presence_penaltydouble or nullOptional-2-2

reasoningobjectOptional

response_formatobjectOptional

seedinteger or nullOptional-9007199254740991-9007199254740991

stopstring or list of strings or nullOptional

streambooleanOptionalDefaults to false

stream_optionsobject or nullOptional

temperaturedouble or nullOptional0-2Defaults to 1

tool_choice"none" or "auto" or "required" or objectOptional

toolslist of objectsOptional

top_pdouble or nullOptional0-1Defaults to 1

debugobjectOptional

Response

Successful chat completion response

idstring

choiceslist of objects

createddouble

modelstring

object"chat.completion"

system_fingerprintstring or null

usageobject or null

1	curl -X POST https://openrouter.ai/api/v1/chat/completions \
2	-H "Authorization: Bearer <token>" \
3	-H "Content-Type: application/json" \
4	-d '{
5	"messages": [
6	{
7	"role": "string",
8	"content": "string"
9	}
10	]
11	}'

1	{
2	"id": "string",
3	"choices": [
4	{
5	"finish_reason": "tool_calls",
6	"index": 1.1,
7	"message": {
8	"role": "string",
9	"content": "string",
10	"name": "string",
11	"tool_calls": [
12	{
13	"id": "string",
14	"type": "string",
15	"function": {
16	"name": "string",
17	"arguments": "string"
18	}
19	}
20	],
21	"refusal": "string",
22	"reasoning": "string"
23	},
24	"logprobs": {
25	"content": [
26	{
27	"token": "string",
28	"logprob": 1.1,
29	"bytes": [
30	1.1
31	],
32	"top_logprobs": [
33	{
34	"token": "string",
35	"logprob": 1.1,
36	"bytes": [
37	1.1
38	]
39	}
40	]
41	}
42	],
43	"refusal": [
44	{
45	"token": "string",
46	"logprob": 1.1,
47	"bytes": [
48	1.1
49	],
50	"top_logprobs": [
51	{
52	"token": "string",
53	"logprob": 1.1,
54	"bytes": [
55	1.1
56	]
57	}
58	]
59	}
60	]
61	}
62	}
63	],
64	"created": 1.1,
65	"model": "string",
66	"object": "string",
67	"system_fingerprint": "string",
68	"usage": {
69	"completion_tokens": 1.1,
70	"prompt_tokens": 1.1,
71	"total_tokens": 1.1,
72	"completion_tokens_details": {
73	"reasoning_tokens": 1.1,
74	"audio_tokens": 1.1,
75	"accepted_prediction_tokens": 1.1,
76	"rejected_prediction_tokens": 1.1
77	},
78	"prompt_tokens_details": {
79	"cached_tokens": 1.1,
80	"audio_tokens": 1.1,
81	"video_tokens": 1.1
82	}
83	}
84	}

Authentication

Request

Response

Errors