Spaces:

---

huggingface

/

HuggingDiscussions

Like 98

Running

Title: Request to register LeastGen as Inference Provider (leastgen)

Hello Hugging Face team,

We are LeastGen (https://leastgen.com) — sovereign GPU network in KSA, RTX A4500 on api.leastgen.com (OpenAI-compatible).

Org: https://huggingface.co/Least-gen (team plan, admin leastgen user hf_OXmc... write token)
Provider slug: leastgen
API base: https://api.leastgen.com (POST /v1/chat/completions, GET /v1/models, POST /billing)
Initial model: Qwen/Qwen3.8-27B → leastgen/Qwen3.8-27B (Q4_K_M 262K vision+tools+reasoning, $0.42/$3.00 per 1M)

Compliance:

• OpenAI-compatible, streaming, usage on both stream/non-stream, Inference-Id header, pricing in /v1/models
• Billing endpoint POST /billing {requestIds:[]} → {requests:[{requestId,costNanoUsd}]} (nano-USD, idempotent)
• pricing.input 0.42 pricing.output 3.00 per 1M, context_length 262144
• datacenters:[{country_code:"SA"}], privacy_policy_url: https://leastgen.com/privacy, terms: https://leastgen.com/terms, zdr:true

JS/Python PRs ready:

• packages/inference/src/providers/leastgen.ts (extends BaseConversationalTask)
• src/huggingface_hub/inference/_providers/leastgen.py
• Will update huggingface/hub-docs with provider page after whitelisting.

Please enable leastgen server-side for POST /api/partners/leastgen/models (currently Cannot POST). We will set mapping to staging first.

Logos SVG attached (light/dark).

Contact: khalid@leastgen.com

Thanks!

Refs: https://huggingface.co/docs/inference-providers/register-as-a-provider