InferIndex
Language, English, EN
Account

ModelsQwenQwen3 8B

Qwen3 8B

qwen/qwen3-8b41k context Compare →
MedianThe median blended price among comparable offers: one price per provider (its cheapest route), standard tier, current prices only. Priority tiers and prices converted from points are not counted.
$0.202
Most expensiveThe highest blended price among comparable offers: one price per provider (its cheapest route), standard tier, current prices only. Priority tiers and prices converted from points are not counted.
$0.47
Offers
5
Providers
5

Cheapest right now

Nscale

via Hugging Face

41k

$0.098 blended / 1M tokens

No change7D
in
$0.07
out
$0.18
Cached input
–
See pricing See pricing at Hugging Face (opens in a new tab)via Hugging Face

Qwen3 8B costs from $0.098 per million tokens (at a 3:1 input:output mix): the cheapest of 3 comparable offers, at Nscale, checked on ; 5 offers are listed.

Listed offers, by price

Comparable offersOther offersSolid steps are comparable offers; dashed steps are the other offers, which are not counted in cheapest, median or most expensive.

Offers not counted in cheapest, median or most expensive (2)

  • $0.123 · Greenference via Eden AI · cannot be designated the cheapest
  • $0.202 · Martian via Martian · cannot be designated the cheapest

Cheapest that confirms no training on your prompts. The provider's published policy: prompts are not used for training.

$0.47 blended / 1M tokens

in
$0.47
out
$0.47
Cached input
–
Go to offer Go to offer at NanoGPT (opens in a new tab)via NanoGPT

Price historyThe cheapest offer each day, against the lab's official list price when it publishes one. Each day at 00:00 UTC, the offer the API would have selected as the cheapest, with the same exclusions as /cheapest. Blended price, promotions included.

Cheapest offer and official list price, by day, 30 days
DayCheapest offerOfficial list price
2026-10-02$0.098, Nscale–
2026-10-03$0.098, Nscale–
2026-10-04$0.098, Nscale–
2026-10-05$0.098, Nscale–
2026-10-06$0.098, Nscale–
2026-10-07$0.098, Nscale–
2026-10-08$0.098, Nscale–
2026-10-09$0.098, Nscale–
2026-10-10$0.098, Nscale–
2026-10-11$0.098, Nscale–
2026-10-12$0.098, Nscale–
2026-10-13$0.098, Nscale–
2026-10-14$0.202, Featherless–
2026-10-15$0.202, Featherless–
2026-10-16$0.098, Nscale–
2026-10-17$0.098, Nscale–
2026-10-18$0.098, Nscale–
2026-10-19$0.098, Nscale–
2026-10-20$0.098, Nscale–
2026-10-21$0.098, Nscale–
2026-10-22$0.098, Nscale–
2026-10-23$0.098, Nscale–
2026-10-24$0.098, Nscale–
2026-10-25$0.098, Nscale–
2026-10-26$0.098, Nscale–
2026-10-27$0.098, Nscale–
2026-10-28$0.098, Nscale–
2026-10-29$0.098, Nscale–
2026-10-30$0.098, Nscale–
2026-10-31$0.098, Nscale–

Cheapest offer

Offers

5 of 5 offers

Offers for Qwen3 8B, one row per offer, sorted by blended price.
RankProvider · viaYou buy access through an aggregator: one account and one API that forward your requests to other providers. "Via" names the aggregator; the price is the one it charges you. When "via" says "hosted API", no aggregator is involved: the model belongs to a lab and another provider sells it.Input $/1MOutput $/1MBlended $/1MA single price that mixes input and output: 3 input tokens for 1 output token, per million tokens. It makes offers easy to rank; your real cost depends on your own mix. ContextThe maximum amount of text, counted in tokens, the model can take in and write in a single request. Past this limit, the request is refused or cut.ConditionsWhat a provider says about how it handles your data and who can sign up: training on prompts, processing region, sign-up requirements. Each comes from the provider's published documents; when nothing is published, it stays unknown.SignalsOffer
1 Nscalevia Hugging Face 0.07  0.18  0.098 41k
See pricing at Hugging Face (opens in a new tab)
2 Greenferencevia Eden AI 0.08  0.25  0.123 41k Unverified. No public page for this offer: listed, not designated cheapestPrice drop observed: The blended price is 73% lower than before this drop; possibly temporary. · Price drop observed since Oct 29, 2026
: Eden AI (opens in a new tab). No public price page for this model
3 Featherless 0.117 0.455 0.202 33k
Go to offer at Featherless (opens in a new tab)
4 Martian 0.117 0.455 0.202 – worldwide Unverified. Neither seller nor price found on its page for 72 hours: listed, not designated cheapest
See pricing at Martian (opens in a new tab)
5 NanoGPT 0.47  0.47  0.47  41k
Go to offer at NanoGPT (opens in a new tab)

No training on promptsPrompts may be used for trainingTraining terms not published

backend not named. The router bills this price itself without naming who serves the model.

Blended = (3 × input + output) / 4, per million tokens. Prices in USD, converted at the ECB rate when the provider bills in another currency.

Method

Monthly cost for your workload

Computed by the API for the three cheapest offers, per 30-day month.
  1. Nscale$6.90per month
  2. Featherless$13.85per month+$6.95 vs the cheapest
  3. NanoGPT$35.25per month+$28.35 vs the cheapest

Host it yourself instead? See the break-even

Compare Qwen3 8B with

Related models

More from Qwen

All Qwen models

Same index family

Use this data

The index is queryable. The API is free to use.
Free API
curl https://api.inferindex.dev/cheapest?model=qwen/qwen3-8b
MCP server
claude mcp add --transport http inferindex https://mcp.inferindex.dev/mcp

Also works in Cursor and Codex: see the MCP page.