Jared Palmer: Kev 4B - API for Egypt, billed in EGP
Livekev-4bKev 4B is a small open-weight decision model from Jared Palmer, built as a LoRA adapter and pointer head on Qwen3.5-4B-Base and served over the same /v1/systemone contract as TypeSafe's
Modality
Decision
Context window
8K tokens
Max output
7K tokens
Region
US
Published
30 Sept 2026
Pricing
Input
2.49
EGP per 1M tokens
Output
Free
EGP per 1M tokens
Monthly cost example
124.47 EGP
Illustrative estimate: 50M input + 15M output tokens per month at this model’s catalog rates.
Use it via the API
Jared Palmer: Kev 4B is a decision model: send the content to judge and typed questions (yes/no, multiple choice, or score) and get one typed answer per question back, with real probabilities. Call it with the OpenAI SDK on /v1/decisions, or with the TypeSafe SDK on /v1/systemone using base URL https://backend.sovereigneg.com. Read the Decisions guide.
Run inference
POST /v1/decisions — OpenAI-compatible Decisions: client.decisions.create in any OpenAI SDK or in seg-sdk. The same model also answers on /v1/systemone for the TypeSafe SDK.
from seg import SEG # pip install seg-sdk
client = SEG(api_key="YOUR_API_KEY")
decision = client.decisions.create(
model="kev-4b",
input="I paid for my order twice and it still has not arrived after a week.",
questions=[
{
"type": "predicate",
"name": "refund",
"instructions": "Is the customer asking for money back?",
},
{
"type": "choice",
"name": "team",
"instructions": "Which team should handle this?",
"choices": [
{"value": "billing", "description": "Charges, refunds, invoices"},
{"value": "shipping", "description": "Delivery and tracking"},
{"value": "technical", "description": "Bugs and outages"},
],
},
{
"type": "score",
"name": "urgency",
"instructions": "How urgent is this?",
"levels": [{"label": "Can wait"}, {"label": "Handle today"}, {"label": "Handle now"}],
},
],
)
refund, team, urgency = decision.answers
print(refund.probability) # probability of yes, 0-1
print(team.choice) # e.g. "billing"
print(urgency.score) # 0-2, can land between levelsfrom openai import OpenAI
client = OpenAI(api_key="YOUR_API_KEY", base_url="https://backend.sovereigneg.com/v1")
decision = client.decisions.create(
model="kev-4b",
input="I paid for my order twice and it still has not arrived after a week.",
questions=[
{
"type": "predicate",
"name": "refund",
"instructions": "Is the customer asking for money back?",
},
{
"type": "choice",
"name": "team",
"instructions": "Which team should handle this?",
"choices": [
{"value": "billing", "description": "Charges, refunds, invoices"},
{"value": "shipping", "description": "Delivery and tracking"},
{"value": "technical", "description": "Bugs and outages"},
],
},
{
"type": "score",
"name": "urgency",
"instructions": "How urgent is this?",
"levels": [{"label": "Can wait"}, {"label": "Handle today"}, {"label": "Handle now"}],
},
],
)
refund, team, urgency = decision.answers
print(refund.probability) # probability of yes, 0-1
print(team.choice) # e.g. "billing"
print(urgency.score) # 0-2, can land between levelsimport OpenAI from "openai";
const client = new OpenAI({ apiKey: "YOUR_API_KEY", baseURL: "https://backend.sovereigneg.com/v1" });
const decision = await client.decisions.create({
model: "kev-4b",
input: "I paid for my order twice and it still has not arrived after a week.",
questions: [
{
type: "predicate",
name: "refund",
instructions: "Is the customer asking for money back?",
},
{
type: "choice",
name: "team",
instructions: "Which team should handle this?",
choices: [
{ value: "billing", description: "Charges, refunds, invoices" },
{ value: "shipping", description: "Delivery and tracking" },
{ value: "technical", description: "Bugs and outages" },
],
},
{
type: "score",
name: "urgency",
instructions: "How urgent is this?",
levels: [{ label: "Can wait" }, { label: "Handle today" }, { label: "Handle now" }],
},
],
});
const [refund, team, urgency] = decision.answers;
console.log(refund.probability); // probability of yes, 0-1
console.log(team.choice); // e.g. "billing"
console.log(urgency.score); // 0-2, can land between levelscurl https://backend.sovereigneg.com/v1/decisions \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "kev-4b",
"input": "I paid for my order twice and it still has not arrived after a week.",
"questions": [
{
"type": "predicate",
"name": "refund",
"instructions": "Is the customer asking for money back?"
},
{
"type": "choice",
"name": "team",
"instructions": "Which team should handle this?",
"choices": [
{ "value": "billing", "description": "Charges, refunds, invoices" },
{ "value": "shipping", "description": "Delivery and tracking" },
{ "value": "technical", "description": "Bugs and outages" }
]
},
{
"type": "score",
"name": "urgency",
"instructions": "How urgent is this?",
"levels": [{ "label": "Can wait" }, { "label": "Handle today" }, { "label": "Handle now" }]
}
]
}'Frequently asked questions
How much does Jared Palmer: Kev 4B cost?
2.49 EGP per 1M input tokens; Output tokens are free. Billing is metered per token in Egyptian pounds (EGP), with no minimum commitment.
What is the context window of Jared Palmer: Kev 4B?
Jared Palmer: Kev 4B supports a context window of 8,192 tokens (8K), with up to 7,372 output tokens per response.
Where is Jared Palmer: Kev 4B hosted?
Jared Palmer: Kev 4B is served from US (US-hosted inference).
How do I use Jared Palmer: Kev 4B via the API?
Jared Palmer: Kev 4B is a decision model. Send the content to judge and typed questions (yes/no, multiple choice, or score) to /v1/decisions with model "kev-4b" and your API key — client.decisions.create in any OpenAI SDK, or SovereignEG's own seg-sdk, with base URL https://backend.sovereigneg.com/v1. The same model also answers on the TypeSafe-compatible /v1/systemone, so the TypeSafe SDK works with base URL https://backend.sovereigneg.com.
Related models
View all models →OpenAI: GPT-6 Luna Decisions
OpenAI · Decision · 1M context
5.93 EGP / 1M input tokens
TypeSafe: Jev 1.13
typesafe · Decision · 262K context
2.49 EGP / 1M input tokens
all-minilm-l12-v2
SBERT · Embedding · 512 context
0.30 EGP / 1M input tokens
all-minilm-l6-v2
SBERT · Embedding · 512 context
0.30 EGP / 1M input tokens