Proprietary model

Fikra 7B V1

fikra-7b-v1

liveproprietary

Fikra 7B V1 costs $0.9 / 1M input and $2.7 / 1M output per 1M tokens, with 32,768 tokens of context and 8,192 max output tokens. After the first-top-up ramp, paid keys run at 500 requests/minute and 50,000 requests/day. Call it at https://api.fikraapi.co.ke/v1 with your fk_live_ key.

Fikra 7B V1 is a proprietary model, fine-tuned from Qwen 2.5 7B Instruct with localization data for Africa.

Published rate

Input
$0.9 / 1M
Cached input
same as input
Output
$2.7 / 1M
Context
32,768 tokens
Max output
8,192 tokens
Rate limit (paid)
500 requests/min · 50,000 requests/day

Prices last changed 2026-11-20. If this page and /v1/models disagree, the API is correct and this page is a bug.

Routing and retention

Model type
Proprietary model fine-tuned by Fikra
Fine-tuned from
Qwen 2.5 7B Instruct
Training data
Localization data for Africa
Upstream family
Fikra
Retention
Billed models are never used for training, even when ZDR is off. ZDR can be enabled per request (X-Fikra-ZDR: true or provider.zdr), and it disables the response cache. Without ZDR, prompts and responses may be cached for up to 300 seconds. Free models collect prompts and responses for training; ZDR does not apply to them. Upstream retention varies by provider.
Call it

OpenAI-compatible. One line changes.

curl https://api.fikraapi.co.ke/v1/chat/completions \
  -H "Authorization: Bearer YOUR_FIKRA_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"fikra-7b-v1","messages":[{"role":"user","content":"Hello"}]}'
from openai import OpenAI

client = OpenAI(base_url="https://api.fikraapi.co.ke/v1", api_key="fk_...")

response = client.chat.completions.create(
    model="fikra-7b-v1",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Try it first

A playground is available for chat models. The quote endpoint returns a price before you run a call.

Open playground

Quote endpoint ->

Need a key? Start with $0.50 credit.

Related: all models · pricing · setup guides · public stats · create an account · support. Operated by Lacesse Ventures. No contractual SLA on the self-serve plan; uptime and incidents are published on /status. Prices and limits as of 2026-11-14.