Chat Completions

!pip install vlmrun

from vlmrun.client import VLMRun

# Initialize the VLM Run client
client = VLMRun(api_key="<VLMRUN_API_KEY>")

# Create a chat completion
response = client.agent.completions.create(
  model="vlmrun-orion-1:auto",
  messages=[{"role": "user", "content": "Who are you and what can you do?"}],
  temperature=0.7,
)

import openai

# Initialize the OpenAI client
client = openai.OpenAI(
    base_url="https://api.vlm.run/v1/openai",
    api_key="<VLMRUN_API_KEY>"
)

# Create a chat completion
response = client.chat.completions.create(
    model="vlmrun-orion-1:auto",
    messages=[{"role": "user", "content": "Who are you and what can you do?"}],
    temperature=0.7,
)

import { VlmRun } from "vlmrun";

// Initialize the VLM Run client
const client = new VlmRun({
  baseURL: "https://api.vlm.run/v1",
  apiKey: "<VLMRUN_API_KEY>"
});

// Create a chat completion
const response = await client.agent.completions.create({
  model: "vlmrun-orion-1:auto",
  messages: [{ role: "user", content: "Who are you and what can you do?" }],
  temperature: 0.7,
});
console.log(response);

{
  "detail": [
    {
      "loc": [
        "<string>"
      ],
      "msg": "<string>",
      "type": "<string>",
      "input": "<unknown>",
      "ctx": {}
    }
  ]
}

POST

openai

chat

completions

!pip install vlmrun

from vlmrun.client import VLMRun

# Initialize the VLM Run client
client = VLMRun(api_key="<VLMRUN_API_KEY>")

# Create a chat completion
response = client.agent.completions.create(
  model="vlmrun-orion-1:auto",
  messages=[{"role": "user", "content": "Who are you and what can you do?"}],
  temperature=0.7,
)

import openai

# Initialize the OpenAI client
client = openai.OpenAI(
    base_url="https://api.vlm.run/v1/openai",
    api_key="<VLMRUN_API_KEY>"
)

# Create a chat completion
response = client.chat.completions.create(
    model="vlmrun-orion-1:auto",
    messages=[{"role": "user", "content": "Who are you and what can you do?"}],
    temperature=0.7,
)

import { VlmRun } from "vlmrun";

// Initialize the VLM Run client
const client = new VlmRun({
  baseURL: "https://api.vlm.run/v1",
  apiKey: "<VLMRUN_API_KEY>"
});

// Create a chat completion
const response = await client.agent.completions.create({
  model: "vlmrun-orion-1:auto",
  messages: [{ role: "user", content: "Who are you and what can you do?" }],
  temperature: 0.7,
});
console.log(response);

{
  "detail": [
    {
      "loc": [
        "<string>"
      ],
      "msg": "<string>",
      "type": "<string>",
      "input": "<unknown>",
      "ctx": {}
    }
  ]
}

!pip install vlmrun

from vlmrun.client import VLMRun

# Initialize the VLM Run client
client = VLMRun(api_key="<VLMRUN_API_KEY>")

# Create a chat completion
response = client.agent.completions.create(
  model="vlmrun-orion-1:auto",
  messages=[{"role": "user", "content": "Who are you and what can you do?"}],
  temperature=0.7,
)

import openai

# Initialize the OpenAI client
client = openai.OpenAI(
    base_url="https://api.vlm.run/v1/openai",
    api_key="<VLMRUN_API_KEY>"
)

# Create a chat completion
response = client.chat.completions.create(
    model="vlmrun-orion-1:auto",
    messages=[{"role": "user", "content": "Who are you and what can you do?"}],
    temperature=0.7,
)

import { VlmRun } from "vlmrun";

// Initialize the VLM Run client
const client = new VlmRun({
  baseURL: "https://api.vlm.run/v1",
  apiKey: "<VLMRUN_API_KEY>"
});

// Create a chat completion
const response = await client.agent.completions.create({
  model: "vlmrun-orion-1:auto",
  messages: [{ role: "user", content: "Who are you and what can you do?" }],
  temperature: 0.7,
});
console.log(response);

Authorizations

Authorization

string

header

required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Headers

user-agent

string | null

Body

application/json

Request payload for the OpenAI chat completions API for VLM Run agents.

messages

Message · object[]

required

Messages to complete

Hide child attributes

messages.role

enum<string>

default:user

Available options:

user,

assistant,

system,

developer,

tool

messages.content

messages.name

string | null

messages.tool_call_id

string | null

messages.tool_calls

ChatCompletionMessageFunctionToolCallParam · object[] | null

Hide child attributes

messages.tool_calls.id

string

required

messages.tool_calls.function

Function · object

required

The function that the model called.

Hide child attributes

messages.tool_calls.function.arguments

string

required

messages.tool_calls.function.name

string

required

messages.tool_calls.type

string

required

Allowed value: "function"

string

ID of the completion

model

VLM Run Agent model to use for completion. When omitted, the skill's vlmrun.yaml model is used; otherwise the agent default.

Available options:

vlmrun-orion-1,

vlmrun-orion-1:auto,

vlmrun-orion-1:fast,

vlmrun-orion-1:pro,

vlmrun-orion-2,

vlmrun-orion-2:auto,

vlmrun-orion-2:qwen3.6-35b-a3b,

vlmrun-orion-2:gemma4-26b-a4b,

vlmrun-orion-2:cosmos3-nano,

vlmrun-orion-2:kimi-2.6,

vlmrun-orion-2:kimi-k3,

vlmrun-orion-2:gpt-5.5,

vlmrun-orion-2:opus-4.8,

vlmrun-orion-2:muse-spark-1.1,

vlmrun-orion-2:grok-4.5,

vlmrun-orion-2:gemini-flash-3.6,

vlmrun-orion-2:fast,

vlmrun-orion-2:pro

max_tokens

integer

default:32768

Maximum number of tokens to generate

integer | null

default:1

Number of completions to generate

temperature

number

default:0

Temperature of the sampling distribution

top_p

number

default:1

Cumulative probability of parameter highest probability vocabulary tokens to keep for nucleus sampling

top_k

integer | null

Number of highest probability vocabulary tokens to keep for top-k-filtering

logprobs

integer | null

Include the log probabilities on the logprobs most likely tokens, as well the chosen tokens

stream

boolean

default:false

Whether to stream the response or not

tools

Tools · object[] | null

OpenAI-compatible client tool definitions. By default (tool_execution=client) the endpoint returns model tool_calls for the client to execute (OpenRouter / OpenAI / pydantic-ai protocol). Set tool_execution=hybrid to also enable server-side tools such as execute_code.

tool_choice

OpenAI-compatible tool_choice (none / auto / required / specific function). Only used when tools is set.

tool_execution

enum<string> | null

How tools are executed when tools is set. client (default): passthrough — all tool_calls returned to the caller. hybrid: server tools (see server_tools) run in-process; only client tool_calls are returned with finish_reason=tool_calls.

Available options:

client,

hybrid

server_tools

string[] | null

Server-owned tool names to inject and execute when tool_execution=hybrid. Defaults to ["execute_code"]. Unknown names are rejected.

preview

Preview card selection for the agent. Accepts: True to enable all known card types, a list[str] of card discriminator names (e.g. ["preview.image", "preview.grid"]) for a focused set, or False/None to disable structured preview output. When enabled, the agent streams <preview data='{...json...}'/> lines through vlmrun.log.print(...) for the chosen card types.

response_format

JSONSchemaResponseFormat · object

Response format for JSON schema mode as per Fireworks AI specification.

JSONSchemaResponseFormat
JSONModeResponseFormat
JSONSchemaResponseFormatStrict

Hide child attributes

response_format.schema

Schema · object

required

JSON schema definition

response_format.type

string

default:json_schema

Allowed value: "json_schema"

service_tier

enum<string> | null

Delivery tier for the request. 'standard'/'default' uses baseline rates, 'flex' applies a 50% discount with higher latency, and 'priority' applies a 1.8x premium for highest reliability. When omitted (or 'auto'), the server default ('standard') applies — opt into 'flex'/'priority' explicitly.

Available options:

auto,

default,

standard,

flex,

priority

session_id

string

Session UUID for persisting the chat history

session_token

string | null

Client-minted session token for IDOR protection (base64url, >=256-bit)

metadata

Metadata · object | null

Additional metadata for the request (e.g., dataset_name, experiment_id, etc.)

skills

AgentSkill · object[] | null

List of agent skills to enable for this request.

Hide child attributes

skills.type

string

default:skill_reference

The type of the skill. Use 'skill_reference' for DB-stored skills referenced by id/name. Use 'inline' to provide the skill as a base64-encoded zip bundle.

skills.skill_id

string | null

The unique identifier of the skill — a UUID or a name string (e.g., 'pillow', 'batch-processing').

skills.skill_name

string | null

Human-readable skill name for lookup (e.g., 'invoice-extraction'). Alternative to skill_id. Deprecated in favour of skill_id.

skills.skill_version

default:latest

The version of the skill — an integer (e.g. 2) or 'latest'.

skills.version

DEPRECATED: Use 'skill_version' instead. The version of the skill.

skills.name

string | null

Human-readable name for the inline skill (used for discovery and logging).

skills.description

string | null

Short description of what the inline skill does.

skills.source

InlineSkillSource · object | null

Source payload for inline skills. Contains the base64-encoded zip bundle with type, media_type, and data fields.

Hide child attributes

skills.source.data

string

required

Base64-encoded zip bundle containing the skill files.

skills.source.type

string

default:base64

Encoding type for the inline skill data. Currently only 'base64' is supported.

Allowed value: "base64"

skills.source.media_type

string

default:application/zip

MIME type of the skill bundle. Must be 'application/zip'.

skills.bundle

string | null

DEPRECATED: Use 'source.data' instead. Base64-encoded zip bundle containing the skill files (inline skills only).

Response

Successful Response

Get Predictions Lookup Agent

Get Started

Health

Generate

Predictions

Chat

Agent

Executions

Files

Models / Hub

Skills

Artifacts

Feedback

Chat Completions

Authorizations

Headers

Body

Response