API Reference

Nova API Reference

Nova provides a unified API for 100+ AI models. Text models use an OpenAI-compatible interface. Image and video models use an async generation pattern — POST to create a job, then poll for the result.

Base URL

https://api.nova.ai

POSTChat Completions

/v1/chat/completions

POSTImage Generation

/api/v1/model/generateImage

POSTVideo Generation

/api/v1/model/generatenvideo

GETPoll Results

/api/v1/model/prediction/{id}

GETList Models

/v1/models

Authentication

All requests require a Nova API key passed as a Bearer token in the Authorization header. Get your key from Console → API Keys.

POSTAuthorization: Bearer <key>

http

Authorization: Bearer nova-YOUR_API_KEY

python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.nova.ai/v1",
    api_key="nova-YOUR_API_KEY",
)

Chat Completions

Fully OpenAI-compatible. Use any OpenAI SDK by pointing base_url at Nova's endpoint. Supports streaming via SSE.

POST/v1/chat/completions

Parameters

modelstring

requiredModel ID in "provider/model" format, e.g. "deepseek/deepseek-r1".

messagesarray

requiredArray of {role, content} objects. Roles: system, user, assistant.

temperaturenumber

Sampling temperature 0–2. Defaults to 1.

max_tokensinteger

Maximum tokens to generate.

streamboolean

Stream via server-sent events. Defaults to false.

top_pnumber

Nucleus sampling probability. Defaults to 1.

python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.nova.ai/v1",
    api_key="nova-YOUR_API_KEY",
)

response = client.chat.completions.create(
    model="deepseek/deepseek-r1",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user",   "content": "What is the capital of France?"},
    ],
    temperature=0.7,
)

print(response.choices[0].message.content)

bash

curl https://api.nova.ai/v1/chat/completions \
  -H "Authorization: Bearer nova-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-r1",
    "messages": [{"role": "user", "content": "Hello!"}],
    "stream": false
  }'

Image Generation

Async text-to-image. POST a job and receive a prediction ID, then poll /prediction/{id} until status is success.

POST/api/v1/model/generateImage

Parameters

modelstring

requiredImage model ID, e.g. "black-forest-labs/flux-dev".

promptstring

requiredText description of the image to generate.

widthinteger

Width in pixels. Defaults to 1024.

heightinteger

Height in pixels. Defaults to 1024.

stepsinteger

Inference steps. Higher = better quality, slower. Defaults to 28.

seedinteger

Random seed for reproducible generations.

image_urlstring

Reference image for image-to-image generation.

python

import requests, time

headers = {"Authorization": "Bearer nova-YOUR_API_KEY"}

resp = requests.post(
    "https://api.nova.ai/api/v1/model/generateImage",
    headers=headers,
    json={
        "model": "black-forest-labs/flux-dev",
        "prompt": "A majestic dragon over snow-capped mountains at sunset",
        "width": 1024,
        "height": 1024,
    }
)
prediction_id = resp.json()["id"]

while True:
    result = requests.get(
        f"https://api.nova.ai/api/v1/model/prediction/{prediction_id}",
        headers=headers,
    ).json()
    if result["status"] == "success":
        print(result["results"][0]["data"][0]["outputs"][0]["url"])
        break
    elif result["status"] == "failed":
        raise Exception("Generation failed")
    time.sleep(2)

Video Generation

Async text-to-video or image-to-video. Poll for completion — videos typically take 30–120 seconds.

POST/api/v1/model/generatenvideo

Parameters

modelstring

requiredVideo model ID, e.g. "kuaishou/kling-16".

promptstring

requiredText description of the video to generate.

durationinteger

Length in seconds. Supported values depend on model: 5 or 10.

image_urlstring

Starting frame for image-to-video generation.

python

import requests, time

headers = {"Authorization": "Bearer nova-YOUR_API_KEY"}

resp = requests.post(
    "https://api.nova.ai/api/v1/model/generatenvideo",
    headers=headers,
    json={
        "model": "kuaishou/kling-16",
        "prompt": "A camera slowly panning across a misty mountain valley at dawn",
        "duration": 5,
    }
)
prediction_id = resp.json()["id"]

while True:
    result = requests.get(
        f"https://api.nova.ai/api/v1/model/prediction/{prediction_id}",
        headers=headers,
    ).json()
    if result["status"] == "success":
        print(result["results"][0]["data"][0]["outputs"][0]["url"])
        break
    time.sleep(5)

Poll Results

Check the status of any async generation job. Returns status and output URLs when complete.

GET/api/v1/model/prediction/{id}

Response fields

idstring

Prediction ID.

statusstring

"generating" | "success" | "failed"

results[].data[].outputs[].urlstring

URL of the generated output. Available for 1–7 days after generation.

json

{
  "id": "pred_01jwxyz",
  "status": "success",
  "results": [
    {
      "data": [
        {
          "outputs": [
            { "url": "https://cdn.nova.ai/outputs/pred_01jwxyz.png" }
          ]
        }
      ]
    }
  ]
}

List Models

Returns all available models on Nova. Response is OpenAI-compatible — model IDs use provider/model-slug format.

GET/v1/models

python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.nova.ai/v1",
    api_key="nova-YOUR_API_KEY",
)

models = client.models.list()
for model in models.data:
    print(model.id)  # e.g. "deepseek/deepseek-r1"

bash

curl https://api.nova.ai/v1/models \
  -H "Authorization: Bearer nova-YOUR_API_KEY"

Errors & Status Codes

Nova uses standard HTTP status codes. Error responses include a message field with details.

400

Bad Request

Invalid parameters or missing required fields.

401

Unauthorized

Missing or invalid API key.

402

Payment Required

Insufficient credit balance. Top up to continue.

404

Not Found

Model or prediction ID does not exist.

429

Rate Limited

Too many requests. Implement exponential backoff and retry.

500

Server Error

Internal error. Retry after a short delay.

API Reference

Nova API Reference

Base URL

https://api.nova.ai

POSTChat Completions

/v1/chat/completions

POSTImage Generation

/api/v1/model/generateImage

POSTVideo Generation

/api/v1/model/generatenvideo

GETPoll Results

/api/v1/model/prediction/{id}

GETList Models

/v1/models

Authentication

All requests require a Nova API key passed as a Bearer token in the Authorization header. Get your key from Console → API Keys.

POSTAuthorization: Bearer <key>

http

Authorization: Bearer nova-YOUR_API_KEY

python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.nova.ai/v1",
    api_key="nova-YOUR_API_KEY",
)

Chat Completions

Fully OpenAI-compatible. Use any OpenAI SDK by pointing base_url at Nova's endpoint. Supports streaming via SSE.

POST/v1/chat/completions

Parameters

modelstring

requiredModel ID in "provider/model" format, e.g. "deepseek/deepseek-r1".

messagesarray

requiredArray of {role, content} objects. Roles: system, user, assistant.

temperaturenumber

Sampling temperature 0–2. Defaults to 1.

max_tokensinteger

Maximum tokens to generate.

streamboolean

Stream via server-sent events. Defaults to false.

top_pnumber

Nucleus sampling probability. Defaults to 1.

python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.nova.ai/v1",
    api_key="nova-YOUR_API_KEY",
)

response = client.chat.completions.create(
    model="deepseek/deepseek-r1",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user",   "content": "What is the capital of France?"},
    ],
    temperature=0.7,
)

print(response.choices[0].message.content)

bash

curl https://api.nova.ai/v1/chat/completions \
  -H "Authorization: Bearer nova-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-r1",
    "messages": [{"role": "user", "content": "Hello!"}],
    "stream": false
  }'

Image Generation

Async text-to-image. POST a job and receive a prediction ID, then poll /prediction/{id} until status is success.

POST/api/v1/model/generateImage

Parameters

modelstring

requiredImage model ID, e.g. "black-forest-labs/flux-dev".

promptstring

requiredText description of the image to generate.

widthinteger

Width in pixels. Defaults to 1024.

heightinteger

Height in pixels. Defaults to 1024.

stepsinteger

Inference steps. Higher = better quality, slower. Defaults to 28.

seedinteger

Random seed for reproducible generations.

image_urlstring

Reference image for image-to-image generation.

python

import requests, time

headers = {"Authorization": "Bearer nova-YOUR_API_KEY"}

resp = requests.post(
    "https://api.nova.ai/api/v1/model/generateImage",
    headers=headers,
    json={
        "model": "black-forest-labs/flux-dev",
        "prompt": "A majestic dragon over snow-capped mountains at sunset",
        "width": 1024,
        "height": 1024,
    }
)
prediction_id = resp.json()["id"]

while True:
    result = requests.get(
        f"https://api.nova.ai/api/v1/model/prediction/{prediction_id}",
        headers=headers,
    ).json()
    if result["status"] == "success":
        print(result["results"][0]["data"][0]["outputs"][0]["url"])
        break
    elif result["status"] == "failed":
        raise Exception("Generation failed")
    time.sleep(2)

Video Generation

Async text-to-video or image-to-video. Poll for completion — videos typically take 30–120 seconds.

POST/api/v1/model/generatenvideo

Parameters

modelstring

requiredVideo model ID, e.g. "kuaishou/kling-16".

promptstring

requiredText description of the video to generate.

durationinteger

Length in seconds. Supported values depend on model: 5 or 10.

image_urlstring

Starting frame for image-to-video generation.

python

import requests, time

headers = {"Authorization": "Bearer nova-YOUR_API_KEY"}

resp = requests.post(
    "https://api.nova.ai/api/v1/model/generatenvideo",
    headers=headers,
    json={
        "model": "kuaishou/kling-16",
        "prompt": "A camera slowly panning across a misty mountain valley at dawn",
        "duration": 5,
    }
)
prediction_id = resp.json()["id"]

while True:
    result = requests.get(
        f"https://api.nova.ai/api/v1/model/prediction/{prediction_id}",
        headers=headers,
    ).json()
    if result["status"] == "success":
        print(result["results"][0]["data"][0]["outputs"][0]["url"])
        break
    time.sleep(5)

Poll Results

Check the status of any async generation job. Returns status and output URLs when complete.

GET/api/v1/model/prediction/{id}

Response fields

idstring

Prediction ID.

statusstring

"generating" | "success" | "failed"

results[].data[].outputs[].urlstring

URL of the generated output. Available for 1–7 days after generation.

json

{
  "id": "pred_01jwxyz",
  "status": "success",
  "results": [
    {
      "data": [
        {
          "outputs": [
            { "url": "https://cdn.nova.ai/outputs/pred_01jwxyz.png" }
          ]
        }
      ]
    }
  ]
}

List Models

Returns all available models on Nova. Response is OpenAI-compatible — model IDs use provider/model-slug format.

GET/v1/models

python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.nova.ai/v1",
    api_key="nova-YOUR_API_KEY",
)

models = client.models.list()
for model in models.data:
    print(model.id)  # e.g. "deepseek/deepseek-r1"

bash

curl https://api.nova.ai/v1/models \
  -H "Authorization: Bearer nova-YOUR_API_KEY"

Errors & Status Codes

Nova uses standard HTTP status codes. Error responses include a message field with details.

400

Bad Request

Invalid parameters or missing required fields.

401

Unauthorized

Missing or invalid API key.

402

Payment Required

Insufficient credit balance. Top up to continue.

404

Not Found

Model or prediction ID does not exist.

429

Rate Limited

Too many requests. Implement exponential backoff and retry.

500

Server Error

Internal error. Retry after a short delay.