Zero data retention

A safer, cheaper, faster inference provider

Zero data retention -- your data is not our data.

Talk to us

or read the quickstart

POST /v1/chat/completionsbase_urlhttps://api.inferway.ai/v1stream true

01Privacy

Private by construction.

No stage on this path persists your data.

Read the transparency hub

Request sent

TLS-encrypted, straight to a dedicated endpoint.

Routed

No shared queue in the path.

Inference in GPU memory

Paged KV, never written to disk.

Destroyed

Erased on completion; only metadata remains.

02Model & pricing

Curated models, honest prices.

from catalog

Qwen3.8 27B

Live

262K ctxNVFP4max out 32,768Zero retention

Per token, metered from metadata. No seats, no minimums.

Get an API key Model detail

Coming soon

Qwen3.8 Flash

inferway/qwen3.8-flash

Coming soon

Qwen's efficiency-focused Flash model.

Text

MiniMax H3

inferway/minimax-h3

Coming soon

MiniMax's general-purpose H3 generation model.

DeepSeek V4 Flash 0731

inferway/deepseek-v4-flash-0731

Coming soon

DeepSeek's efficiency-focused V4 Flash release.

Text

GLM-5.3 Flash

inferway/glm-5.3-flash

Coming soon

Z.ai's Flash model for coding and agentic workloads.

03Benchmarks

Numbers are being re-measured.

Measuring now

First-token latencypendingMeasured at fixed concurrency

Output throughputpendingTokens per second, single stream

90-day uptimependingPer service, from the status page

Throughput and latency are being re-measured on the current hardware and model. Until that finishes, this section stays empty — we publish measurements with hardware, concurrency and timestamp attached, or we publish nothing.

04Integrate

Keep your SDK.

Create an accountConsole

no credit card

Point base_url at Inferway

https://api.inferway.ai/v1

Run the streaming smoke test

stream=true

Free use: 60/min · 300/hour · 1000/day

Long outputs require stream=true

Streaming

curl https://api.inferway.ai/v1/chat/completions \
  -H "Authorization: Bearer $INFERWAY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "inferway/qwen3.8-27b",
    "stream": true,
    "messages": [{"role": "user", "content": "Say hello in one short sentence."}]
  }'

Copy setup for your AI assistant

Paste into Claude Code or Cursor. Key stays in your .env, never in the prompt.

copy

05Playground

Try it without an account.

No key needed · free use: 60/min · 300/hour · 1000/day

No guest model is available right now

Ask about data retention, pricing, streaming, or limits — or just chat.

Checking live model availability…

Questions

No. Content is never written to disk — inference runs entirely in GPU memory and is destroyed on completion. We retain request metadata only (token counts, timestamps, latency, routing, HTTP status) for up to 90 days, for billing and reliability.

Input tokens that hit the cache are billed at the cache rate rather than the normal input rate, so long shared system prompts and retrieval contexts benefit most. All three rates are the ones shown above, published for the market you are being served.

Non-streaming requests pass through a CDN with roughly a 100-second first-byte timeout. A long output returns no first byte until generation finishes, so it hits that timeout. With stream=true, bytes start arriving with the first token. Non-streaming is fine for short completions.

You pay per token, metered from request metadata, never from your request content. Pricing and context limits are published on the model catalog and generated from a single catalog source.

Longer answers: Docs · Privacy · Transparency

Start with a key.

Read the quickstart