Cheapest open-weight models.
prices as of · re-ranked daily · 149 qualifying models
Anyone may host an open-weight model, so several providers compete on the same weights and the spread between them is wide. 149 of them are sold by 3 providers or more, from $0.04. Prices are per million tokens, three parts input to one part output, at each provider's standard rate.
Today's pick is Llama 3.2 1B Instruct at Inference, $0.01 in and $0.01 out per million tokens, read on 16 Sept 2026.
- 1
InferenceLlama 3.2 1B Instruct
open weights, 60k context, sold by 11 providers, $0.01 / $0.01 per M tokens at InferenceHosts the open weights on its own servers, so the price is its own, not a resale. Precision not disclosed.
$0.04/M tokens
- 2Inference
Llama 3.2 3B Instruct
open weights, 131k context, sold by 17 providers, $0.02 / $0.02 per M tokens at InferenceHosts the open weights on its own servers, so the price is its own, not a resale. Precision not disclosed.
$0.08/M tokens
- 3
Nous PortalMistral Nemo
open weights, 128k context, sold by 25 providers, $0.02 / $0.03 per M tokens at Nous PortalResells OpenRouter's catalog at 0.80× the listed price, funded by a subscription whose credits carry a 10% bonus.
$0.08/M tokens
- 4
Merge GatewayGLM-5.3-Flash
open weights, 1M context, sold by 76 providers, $0.01 / $0.05 per M tokens at Merge GatewayHosts the open weights on its own servers, so the price is its own, not a resale. Precision not disclosed.
$0.10/M tokens
- 5
Kilo GatewayLlama 3.1 8B Instruct
open weights, 131k context, sold by 24 providers, $0.02 / $0.04 per M tokens at Kilo GatewayRoutes to a host serving the open weights; precision not disclosed.
$0.10/M tokens
- 6
OpenRouterLing 3.0 Flash
open weights, 262k context, sold by 12 providers, $0.02 / $0.06 per M tokens at OpenRouterThe same price OpenRouter charges.
$0.13/M tokens
- 7Kilo Gateway
gpt-oss-20b
open weights, 131k context, sold by 43 providers, $0.02 / $0.10 per M tokens at Kilo GatewayRoutes to a host serving the open weights; precision not disclosed.
$0.16/M tokens
- 8
Azure Cognitive ServicesMinistral 3B (latest)
open weights, 128k context, sold by 14 providers, $0.04 / $0.04 per M tokens at Azure Cognitive ServicesThe same price Mistral AI charges.
$0.16/M tokens
- 9
NanoGPTNemotron 3.5 Lightning
open weights, 262k context, sold by 12 providers, $0.05 / $0.01 per M tokens at NanoGPTRoutes to a host serving the open weights; precision not disclosed.
$0.16/M tokens
- 10Merge Gateway
Gemma 3 4B
open weights, 131k context, sold by 10 providers, $0.04 / $0.08 per M tokens at Merge GatewayHosts the open weights on its own servers, so the price is its own, not a resale. Precision not disclosed.
$0.20/M tokens
Frequently asked
- What is the cheapest of the open-weight models right now?
- Llama 3.2 1B Instruct at Inference, $0.01 in and $0.01 out per million tokens, read on 16 Sept 2026.
- Which lab shows up most in this ranking?
- Meta trained 3 of the top 10 models here.
- How is this ranked, and how current is it?
- Every row is the cheapest price for that model that we can account for: the maker's own, a price below it with a published reason, or a price at or above it. A price below the maker's own with nothing to explain it is shown on the model's page and ranked nowhere. Prices are read daily from 212 providers' public catalogs and blended three parts input to one part output per million tokens. Last refresh: 16 Sept 2026.