Create chat completion

curl --request POST \
  --url https://api.rav.chat/v1/chat/completions \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "ravchat",
  "messages": [
    {
      "role": "user",
      "content": "What does the Tanya say about simcha?"
    }
  ]
}
'

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1720512000,
  "model": "ravchat",
  "choices": [
    {
      "index": 123,
      "message": {
        "role": "<string>",
        "content": "<string>",
        "refusal": null
      },
      "logprobs": null
    }
  ],
  "usage": {
    "prompt_tokens": 123,
    "completion_tokens": 123,
    "total_tokens": 123,
    "cost": "2.400"
  },
  "x_ravchat": {
    "project_id": "<string>",
    "workspace_id": "<string>",
    "session_id": "<string>",
    "thread_id": "<string>",
    "created": {
      "project": true,
      "workspace": true,
      "session": true
    },
    "credits": "<string>",
    "files": [
      {
        "id": "file-rav-0f8ff0ae",
        "object": "file",
        "bytes": 4812,
        "created_at": 1720512000,
        "filename": "summary.md",
        "purpose": "assistants_output",
        "status": "processed",
        "project_id": "<string>",
        "workspace_id": "<string>",
        "path": "summary.md",
        "session_id": "<string>"
      }
    ]
  }
}

{
  "error": {
    "message": "n must be 1",
    "type": "invalid_request_error",
    "param": "n",
    "code": null
  }
}

{
  "error": {
    "message": "Invalid API key",
    "type": "invalid_request_error",
    "param": null,
    "code": "invalid_api_key"
  }
}

{
  "error": {
    "message": "Missing API key scope",
    "type": "permission_error",
    "param": null,
    "code": "missing_scope"
  }
}

{
  "error": {
    "message": "The model `gpt-4` does not exist",
    "type": "invalid_request_error",
    "param": null,
    "code": "model_not_found"
  }
}

{
  "error": {
    "message": "Idempotency-Key was reused with a different request body",
    "type": "invalid_request_error",
    "param": null,
    "code": "idempotency_key_reused"
  }
}

{
  "error": {
    "message": "You exceeded your current quota",
    "type": "insufficient_quota",
    "param": null,
    "code": "insufficient_quota"
  }
}

POST

chat

completions

curl --request POST \
  --url https://api.rav.chat/v1/chat/completions \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "ravchat",
  "messages": [
    {
      "role": "user",
      "content": "What does the Tanya say about simcha?"
    }
  ]
}
'

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1720512000,
  "model": "ravchat",
  "choices": [
    {
      "index": 123,
      "message": {
        "role": "<string>",
        "content": "<string>",
        "refusal": null
      },
      "logprobs": null
    }
  ],
  "usage": {
    "prompt_tokens": 123,
    "completion_tokens": 123,
    "total_tokens": 123,
    "cost": "2.400"
  },
  "x_ravchat": {
    "project_id": "<string>",
    "workspace_id": "<string>",
    "session_id": "<string>",
    "thread_id": "<string>",
    "created": {
      "project": true,
      "workspace": true,
      "session": true
    },
    "credits": "<string>",
    "files": [
      {
        "id": "file-rav-0f8ff0ae",
        "object": "file",
        "bytes": 4812,
        "created_at": 1720512000,
        "filename": "summary.md",
        "purpose": "assistants_output",
        "status": "processed",
        "project_id": "<string>",
        "workspace_id": "<string>",
        "path": "summary.md",
        "session_id": "<string>"
      }
    ]
  }
}

{
  "error": {
    "message": "n must be 1",
    "type": "invalid_request_error",
    "param": "n",
    "code": null
  }
}

{
  "error": {
    "message": "Invalid API key",
    "type": "invalid_request_error",
    "param": null,
    "code": "invalid_api_key"
  }
}

{
  "error": {
    "message": "Missing API key scope",
    "type": "permission_error",
    "param": null,
    "code": "missing_scope"
  }
}

{
  "error": {
    "message": "The model `gpt-4` does not exist",
    "type": "invalid_request_error",
    "param": null,
    "code": "model_not_found"
  }
}

{
  "error": {
    "message": "Idempotency-Key was reused with a different request body",
    "type": "invalid_request_error",
    "param": null,
    "code": "idempotency_key_reused"
  }
}

{
  "error": {
    "message": "You exceeded your current quota",
    "type": "insufficient_quota",
    "param": null,
    "code": "insufficient_quota"
  }
}

Authorizations

Authorization

string

header

required

A RavChat API key: Authorization: Bearer sk-rav-...

Body

application/json

model

string

required

A RavChat model id. See the Models catalog.

Example:

"ravchat"

messages

object[]

required

The conversation. system/developer become agent context; the final user message is the query.

Show child attributes

stream

boolean

default:false

Stream chat.completion.chunk SSE. Strongly recommended — turns can take minutes.

stream_options

object

Show child attributes

response_format

Text · object

text, json_object, or a strict json_schema. Schema is enforced natively; non-conforming output is charged, not retried.

Text
JSON object
JSON schema

Show child attributes

metadata

object

metadata.ravchat.mode = "fast" selects fast mode (equivalent to the top-level mode field).

Show child attributes

max_tokens

integer

Output budget hint where the runtime can honor it.

integer

default:1

Must be 1.

stop

Accepted.

temperature

number

Accepted and ignored — the agent controls generation.

user

string

Accepted, logged for abuse attribution only.

mode

enum<string>

default:standard

Answering mode. fast runs the quick-answer lane: the agent performs one bounded source search (the local Chabad corpus index, or a direct reference lookup) and returns a concise, sourced answer in your language — same guardrails (verified citations, hashkafa frame, never a halachic ruling), a fraction of the latency. standard (default) runs the full study agent. Also selectable via fast: true or metadata.ravchat.mode = "fast" (the metadata form passes through every OpenAI SDK without extra_body).

Available options:

standard,

fast

fast

boolean

default:false

Alias for mode: "fast". If both are present, mode wins.

Response

A chat completion. When stream:true, a text/event-stream of chat.completion.chunk objects terminated by data: [DONE].

string

Example:

"chatcmpl-abc123"

object

string

Example:

"chat.completion"

created

integer

Example:

1720512000

model

string

Example:

"ravchat"

choices

object[]

Show child attributes

usage

object

Show child attributes

x_ravchat

object

RavChat extension — where the turn ran and what it produced. Also mirrored in x-ravchat-* response headers.

Show child attributes

List models