> Nscale
Nscale's Serverless Inference is an OpenAI-compatible endpoint service for open models - the catalogue pages list Meta Llama 4 Scout, Alibaba Qwen, DeepSeek, OpenAI GPT OSS 20B/120B, Mistral 8x22B Instruct, FLUX.1 [schnell] and Stable Diffusion XL Base 1.0 - with token-based pay-as-you-go billing. Per Nscale's own launch materials, pricing is based on input and output tokens for text and vision models, while image generation is priced on output image dimensions. Per-token rates are not published on a static official page: the docs direct users to the console's AI Services to Models list or the /v1/models API endpoint for pricing and context length. Every new user can claim $5 of free credit, and Nscale states the platform does not log or train on request or response content.
// pros
- New users can claim $5 of free credit before adding a payment card, which is enough to exercise the OpenAI-compatible endpoints and image generation at small volume.
- Published data-use commitment on the launch post: the Serverless Inference platform "does not log or train on request or response content", stated as a privacy guarantee rather than an opt-in.
- OpenAI API and SDK compatibility is stated explicitly, so existing code and tooling can be repointed at Nscale with minimal change.
- Model pricing and context length are retrievable programmatically from the /v1/models endpoint, which allows automated price comparison without scraping a marketing page.
- Fine-tuning on enterprise data is offered in the same console with a low stated entry threshold - "can be started by any developer with $2 credit" - and a workbench for side-by-side prompt and parameter comparisons.
// cons
- No static per-1M-token rate card is published: docs redirect to the console model list or the /v1/models API, so pre-costing requires an account or an API call.
- nscale.com/pricing returns a 404, and the "pricing page" link in the launch post now redirects to the AI Services product overview rather than a price table.
- The $5 free-credit terms are not itemised - expiry, whether the credit applies to fine-tuning as well as inference, and any minimum spend are not stated on the pages fetched.
- Positioning is enterprise and sovereign-infrastructure oriented (60MW renewable data centre in Norway, 1.3GW pipeline, thousands-of-GPUs sales motion), so small self-serve workloads are not the target profile.
Nscale is worth evaluating when European data boundaries and token-billed open-model inference both matter, and the $5 credit makes a first pass cheap; because rates are not on a static page, query /v1/models or the console first, then compare against a token provider with a published rate card.
written by Mike Fleming, Uprouter editorial · verified against official sources September 16, 2026
- Nscale launch post - Serverless Inference, token-based pricing and $5 free credit ↗
- Nscale press release - Serverless Inference platform, token-based pay-as-you-go ↗
- Nscale AI Services (inference endpoints, fine-tuning, $2 credit fine-tuning threshold) ↗
- Nscale docs - model list location and /v1/models pricing endpoint ↗
How we estimated this: the free allowance is converted to a dollar value using the provider’s own pay-as-you-go rates for the models a typical developer would reach for. When the allowance is metered in requests, tokens or minutes rather than dollars, we assume median usage of the cheapest capable model. Because assumptions are involved, the figure carries the est. flag — treat it as an order of magnitude, not a quote.
note: Every new user can claim $5 of free credit, stated in Nscale's Serverless Inference launch post (2 April 2025) and in its announcement that pricing is based on input and output tokens for text and vision models.
| Plan | Type | Monthly | Input /1M | Output /1M | Note |
|---|---|---|---|---|---|
| Free tier (source-reported) | free | — | — | — | New users get a one-time $5 free credit to explore the models; after that you add a card to buy more. The core is pay-as-you-go (e.g. GPT-OSS 120B ~$0.1 in / $0.4 out per 1M tokens), with no rate limits and no cold starts. |
| Pay-as-you-go | payg | — | — | — | Token-based pricing for serverless inference. |
Pricing normalized from public sources — always verify with the provider.
// model prices
| Model | Input /1M | Output /1M | Context |
|---|---|---|---|
| Llama family | — | — | — |
| DeepSeek family | — | — | — |
| gpt-oss-120b | — | — | 131k |
$ Does Nscale have a free tier?
Yes — Nscale offers a free tier that we value at roughly $5.00 in free credits (estimate — derived from rate-limited usage at pay-as-you-go prices). Always confirm current limits on Nscale's official pricing page — free tiers change without notice.
$ How much does Nscale cost per 1M tokens?
Nscale does not publish simple per-1M-token pricing (it may bill per request, per output, or via a subscription plan). Check their official pricing page for exact figures.
$ Is Nscale safe to use?
Uprouter rates Nscale as low risk. Risk is moderate. The commercial model is token-based pay-as-you-go with a $5 entry credit, and the key policy claim - that Nscale does not log or train on request or response content - is stated by Nscale itself in the launch post, which is a favourable data-use position. Against that, the pricing surface is unstable (nscale.com/pricing is a 404 and per-model token rates are not published statically), the free-credit terms are undated beyond April 2025, and capacity commitments are aimed at reserved enterprise clusters, so published terms should be re-verified at sign-up. Risk ratings are editorial, evidence-based, and never influenced by affiliate relationships.
Raw community sentiment, shown unblended. Votes never feed into objective data fields like price, risk, or status — and never into sort order.
No published reviews yet — be the first.
// more_informationprice history · live status · risk & confidence · code snippets
price_history
No recorded changes yet.
live_status
full history →best-effort probes · not a guarantee · last 15h ago
| Checked | Status | Latency |
|---|---|---|
| 15h ago | up | 280ms |
| 1d ago | up | 278ms |
| 3d ago | up | 278ms |
| 4d ago | up | 283ms |
| 6d ago | up | 270ms |
| 7d ago | up | 273ms |
| 9d ago | up | 286ms |
| 10d ago | up | 281ms |
| 12d ago | up | 282ms |
| 13d ago | up | 289ms |
| 14d ago | up | 277ms |
| 15d ago | up | 277ms |
| 15d ago | up | 285ms |
| 16d ago | up | 268ms |
| 17d ago | up | 275ms |
| 18d ago | up | 283ms |
| 18d ago | up | 288ms |
| 19d ago | up | 281ms |
| 20d ago | up | 284ms |
| 21d ago | up | 280ms |
| 22d ago | up | 270ms |
| 22d ago | up | 268ms |
| 23d ago | up | 281ms |
| 24d ago | up | 282ms |
| 25d ago | up | 279ms |
| 26d ago | up | 281ms |
| 26d ago | up | 280ms |
| 27d ago | up | 286ms |
| 28d ago | up | 271ms |
| 29d ago | up | 281ms |
Uprouter runs its own probes; best-effort, not a guarantee. Probe results depend on our network location and cadence — for production, always use the provider’s own status page.
risk_&_confidence
Risk is moderate. The commercial model is token-based pay-as-you-go with a $5 entry credit, and the key policy claim - that Nscale does not log or train on request or response content - is stated by Nscale itself in the launch post, which is a favourable data-use position. Against that, the pricing surface is unstable (nscale.com/pricing is a 404 and per-model token rates are not published statically), the free-credit terms are undated beyond April 2025, and capacity commitments are aimed at reserved enterprise clusters, so published terms should be re-verified at sign-up.
Confidence reflects how much verifiable evidence (official docs, probe history, community reports) backs this entry. Risk is our editorial assessment of operator reliability and terms — not financial advice.
code_snippets
# Nscale's own API — get a key from their dashboard
curl https://YOUR-PROVIDER-ENDPOINT/chat/completions \
-H "Authorization: Bearer THEIR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "llama-family",
"messages": [{ "role": "user", "content": "Hello" }]
}'Direct to the provider — Uprouter receives nothing; verify their current pricing before purchasing.
Ranked by neutral similarity signals only — live status, free-tier shape and free-credit proximity. Commercial relationships never influence this list.