Providers /InferenceNet
Inference provider

InferenceNet

Models
6
Data residency
No answer on record
00

Our take

Editorial by LLMap · updated Sep 14, 2026

InferenceNet is a boutique hosted inference provider with a tightly curated catalogue of three offers and a verified zero-retention, no-training data policy. It suits low-cost experimentation on its in-house Schematron v2 family or high-throughput access to Kimi K3.

Use this for low-cost experimentation on the Schematron v2 family, where the turbo variant is the cheapest tracked offer. Pick it when zero data retention with an explicit no-training guarantee is required, or for high-throughput needs on Kimi K3 at 49 tokens per second if speed matters more than per-token cost. Skip it if you need a broad model catalogue, verified SOC 2, a confirmed EU endpoint, or a known headquarters location.

Strengths
  • Cheapest entry point among tracked Schematron v2 variants — the turbo variant undercuts its small sibling by roughly two-fifths on input and more than a third on output.
  • Explicit zero-retention, no-training data policy: both are verified rather than merely claimed.
  • Kimi K3 throughput dramatically exceeds in-house models: 49 tokens per second, seven times the turbo variant and almost twenty-five times the small variant.
Trade-offs
  • Extremely limited catalogue depth — only three tracked offers versus broader catalogues elsewhere.
  • Schematron v2-small is slower and more expensive than its turbo sibling on both price and speed: two-thirds higher input price, roughly half again higher output price, and three-and-a-half times slower throughput.
  • Compliance and geographic transparency gaps: SOC 2, headquarters country, and EU endpoint are all unverified in our data.
01

Point your tools here

We have not recorded what a router needs for InferenceNet yet — no base URL and no docs link on file. That is our gap, not a sign InferenceNet has no API; its own documentation is the place to look until we close it.

02

Privacy & data handling

These answers cover requests sent to InferenceNet through OpenRouter, as OpenRouter records them, last read 58 min ago. For InferenceNet's own API we hold no answer: check its terms. SOC 2 and the data processing agreement below are hand-checked, not part of that sync.

Trains on prompts
✗No
Logs prompts
✗No
✓Offered
No answer on record
No answer on record

Prompt retention: none.

03

Models & pricing

Something wrong on this page? Tell us