Overview
Fikra API is an OpenAI-compatible AI inference API for developers and growing AI teams. It exposes chat completions and embeddings endpoints behind a single OpenAI-style base URL. Fikra API does not list a dedicated image generation endpoint. To generate images, use a chat model to return image source code such as SVG, or to produce a detailed prompt that you can pass to an image rendering service.
Prerequisites
- A Fikra API key. Keys are scoped and revocable.
- Set the key as the
FIKRA_API_KEYenvironment variable. - Python with the
openaiSDK installed. - Use the API base URL
https://api.fikraapi.co.ke/v1.
Step-by-step
- Install the OpenAI SDK:
pip install openai
- Create a chat completion that asks the model to generate an SVG image.
import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.fikraapi.co.ke/v1",
api_key=os.environ["FIKRA_API_KEY"],
)
response = client.chat.completions.create(
model="deepseek-v4-flash-0731",
messages=[
{"role": "user", "content": "Generate an SVG image of a sunset over Nairobi skyline."}
],
)
svg = response.choices[0].message.content
print(svg)
- Save the returned SVG string to a file and open it in a browser or image viewer.
Verify it works
Check that response.choices[0].message.content contains valid SVG markup, usually starting with <svg. The response follows the OpenAI chat-completions format, so standard OpenAI SDK handling applies.
Costs
Fikra API charges per 1M tokens for chat models. For example, deepseek-v4-flash-0731 costs 0.09 USD per 1M input tokens and 0.15 USD per 1M output tokens. Other models have different rates; see the pricing page for the full list.
Common errors
- 401: bad or inactive API key. Check that
FIKRA_API_KEYis set correctly. - 400: unsupported model. Use a model listed in the Fikra API documentation.
- 429: rate limit exceeded. Unverified keys get 30 requests/min; trusted keys get 100 requests/min.
- 503: upstream busy. Retry with exponential backoff.