Nscale API pricing
Updated
Nscale lists standard per-token prices on its own pricing page for 4 models. The cheapest is Qwen3.8-Flash-Next-FP8 at USD 0.15 input and USD 0.47 output per 1M tokens; the dearest is GLM 5.3 at USD 1.40 / USD 4.40. It publishes no cached-input prices, and lists no per-model batch, flex or priority prices on the pages we read. Checked 2026-11-15.
- Models with a standard price
- 4
- Median input / output
- USD 0.60 / USD 2.50
- Cached-input prices
- 0 of 4
- Cheapest host for
- 0 of 3 shared models
Which models is Nscale the cheapest provider for?
None. Every one of the 3 models that Nscale shares with other providers is cheaper elsewhere at the standard tier. The closest is Mistral Large 3, where Nscale ranks 3 of 3. It is the dearest provider for 1.
How does Nscale price the most widely hosted models?
The 3 models Nscale sells that the most other providers also sell, with Nscale's rank among them on the 3:1 blend.
| Model | Nscale | Rank | Cheapest |
|---|---|---|---|
| GLM 5.3 | USD 1.40 / USD 4.40 | 6 of 24 | DeepInfra USD 0.5625 / USD 2.50 |
| Qwen3.8 27B | USD 0.20 / USD 2.50 | 6 of 15 | Chutes USD 0.039 / USD 1.25 |
| Mistral Large 3 | USD 1 / USD 2.50 | 3 of 3 | Amazon Bedrock USD 0.50 / USD 1.50 |
Which model makers does Nscale sell?
Nscale prices models from 3 makers. Qwen leads with 2 models. Standard prices, USD per 1M input / output tokens.
| Maker | Models | Cheapest model | Median input / output |
|---|---|---|---|
| Qwen | 2 | Qwen3.8-Flash-Next-FP8 USD 0.15 / USD 0.47 | USD 0.175 / USD 1.485 |
| Mistral AI | 1 | Mistral Large 3 USD 1 / USD 2.50 | — |
| Z.ai | 1 | GLM 5.3 USD 1.40 / USD 4.40 | — |
Nscale price list
USD per 1M tokens: 4 prices for 4 models, one row per model with any other service tiers stacked under the standard price. Sources checked 2026-11-15 unless a row says otherwise. "vs maker" compares a standard price with the model maker's own API on the 3:1 blend; a minus means Nscale is cheaper. Of 2 hosted model prices with a maker price to compare, 1 is below the maker's and 0 above.
| Model | Tier | Input | Cached | Output | Context | vs maker | Source |
|---|---|---|---|---|---|---|---|
| GLM 5.3 | standard | USD 1.40 | — | USD 4.40 | not stated | same | source |
| Mistral Large 3 | standard | USD 1 | — | USD 2.50 | not stated | — | source |
| Qwen3.8 27B | standard | USD 0.20 | — | USD 2.50 | not stated | −31% | source |
| Qwen3.8-Flash-Next-FP8 | standard | USD 0.15 | — | USD 0.47 | not stated | — | source |
- Cached "—" means Nscale's page shows no cache price for that row; budget cached tokens at the input price.
- Context "not stated": Nscale's pricing page does not give context windows.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What service tiers does Nscale publish?
Standard prices only: Nscale publishes no per-model batch, flex, priority or fast price on the pages we read.
Cached input: no cache-read price is published, so repeated prompts bill at the full input price.
What we recorded about Nscale's pricing page: Nscale's website has no public pricing page (nscale.com/pricing returns 404); prices come from the "Available models" table in Nscale's serverless docs ("Prices are in USD per 1 million tokens"). No cached-input price or context length is published there. Qwen3-Embedding-8B excluded (embeddings).
Is Nscale cheaper than the model makers' own APIs?
Mostly yes. Of 2 standard prices for models whose maker sells its own API, 1 is below the maker's price, 0 above and 1 the same, on the 3:1 blend. The median difference is 16% below the maker. The biggest saving is Qwen3.8 27B, 31% below Alibaba Cloud.
BroadNet Inference and Nscale
BroadRouter is BroadNet's own site, so BroadNet is never ranked on this page. BroadNet Inference also serves 2 models that Nscale prices: Qwen3.8 27B at USD 0.175 / USD 2.17 against Nscale's USD 0.20 / USD 2.50 (BroadNet is cheaper on the 3:1 blend); GLM 5.3 at USD 0.78 / USD 3.46 against Nscale's USD 1.40 / USD 4.40 (BroadNet is cheaper on the 3:1 blend). BroadNet prices are from its rate card, synced 2026-11-14. BroadNet Inference pricing.
More on Nscale
- Nscale's own pricing page (the source; checked 2026-11-15)
- All 41 providers compared · Cheapest LLM APIs right now · Download every price (CSV, JSON)
Questions
How much does the Nscale API cost per token?
Nscale's standard prices run from USD 0.15 input and USD 0.47 output per 1M tokens (Qwen3.8-Flash-Next-FP8) to USD 1.40 / USD 4.40 (GLM 5.3). The median across 4 standard prices is USD 0.60 input and USD 2.50 output per 1M, which is USD 0.0000006 per input token. Input, cached input and output are billed separately.
What is the cheapest model on Nscale?
Qwen3.8-Flash-Next-FP8, at USD 0.15 input and USD 0.47 output per 1M tokens (USD 0.23 on a 3:1 input-to-output blend). The cheapest from other makers: Mistral Large 3 (Mistral AI) at USD 1 / USD 2.50; GLM 5.3 (Z.ai) at USD 1.40 / USD 4.40.
Is Nscale cheaper than other providers?
It depends on the model. Of 3 models that other providers also price, Nscale is the cheapest standard price for none; another provider is cheaper for 3; and it is the dearest for 1. Its median rank is 6 of a median 15 providers, on a 3:1 input-to-output blend.
Does Nscale have batch, cached or priority pricing?
Cached input: no cached price is published on its pricing page, so budget repeated prompts at the input price. No batch, flex or priority prices are listed per model on the pages we read.
How much does a chatbot cost per month on Nscale?
For 2,000 requests a day, 1,500 input tokens (60% cached), 300 output, over 30 days: USD 21.96 on Qwen3.8-Flash-Next-FP8, USD 63.00 on Qwen3.8 27B (the middle of Nscale's list) and USD 205.20 on GLM 5.3. Try your own numbers.