Offerings

Every active offering: a model as served by one host, with its current canonical price. Sort by clicking a column header; filter with the row below. Prices are USD per 1M tokens unless the unit toggle is set to per 1M characters (an approximation using a measured, blended tokens-per-character ratio).

Showing every active offering, including the long tail. Show curated models only

More columns: + Vendor + Region - Cache + Batch + Tokenizer + Last change + Source All columns Default columns

ModelHostVariantRegionInputOutputBlended ↑?Cache readCache writeContextConfidence?
Gemini 2.5 Flash-LiteGoogleflex–0.05000.2000.08750.0050000.04171,048,576api
Phi-4 3 sourcesDeepInfra––0.07000.1400.0875––16,384official
MythoMax 13BParasail––0.08000.1100.0875––8,192api
Qwen Turbo 2 sourcesAlibaba Cloud––0.05000.2000.0875––129,024aggregator
qwen-turbo-latestAlibaba Cloud––0.05000.2000.0875––1,000,000aggregator
Nemotron 3.5 Lightning 30B A3B 2 sourcesFireworks AI––0.05000.2000.08750.0100–262,144aggregator
Nemotron 3 Nano 30B A3B 2 sourcesNovita AI––0.05000.2000.0875––262,144aggregator
Nemotron 3 Nano 30B A3B 2 sourcesDeepInfra––0.05000.2000.08750.0250–262,144aggregator
gpt-oss-20bTogether AI––0.05000.2000.0875––131,072aggregator
MiniMax M3.1 FlashTenstorrentfp8–0.05000.2000.08750.005000–1,048,576api
Ling 3.0 Flash VLDeepInfra––0.06000.1800.09000.0120–131,072api
Ling 3.0 Flash FinDeepInfrafp4–0.06000.1800.09000.0120–262,144api
Ling 3.0 Flash 3 sourcesNovita AI promo −65%––0.06000.1800.09000.0120–262,144api
Ling 3.0 Flash 2 sourcesDeepInfra––0.06000.1800.09000.0120–262,144api
gpt-oss-20b 2 sourcesAWS Bedrock–eu-west-10.07000.1500.0900––131,072api
gpt-oss-20b 3 sourcesAWS Bedrock––0.07000.1500.0900––131,072api
MythoMax 13B 2 sourcesNovita AI––0.09000.09000.0900––8,192aggregator
inclusionai/ling-3.0-flash-fastNovita AI––0.06000.1800.09000.0120–262,144aggregator
Embed v4Azure AI Foundry––0.120free0.0900––128,000aggregator
DeepSeek-V4-Flash-0731DeepInfra––0.06000.1800.09000.0150–1,048,576official
DeepSeek-V4-Flash-0731DeepInfrafp8–0.06000.1800.09000.0150–1,048,576api
Ling 3.0 Flash VLDeepInfrafp4–0.06000.1800.09000.0120–131,072api
Qwen3.5-9BDeepInfraflex–0.08000.1200.0900––262,144api
Ling 3.0 FlashDeepInfrafp4–0.06000.1800.09000.0120–262,144api
DeepSeek V4 FlashDeepInfraflex+fp8–0.07200.1440.09000.0144–1,048,576api
Qwen3.5-9BDarkbloomfp4–0.08000.1300.09250.0400–262,144api
Ternary Bonsai 2 9BDarkbloomint4–0.08000.1300.09250.0400–262,144api
Ternary Bonsai 2 9BOpenRouter––0.08000.1300.09250.0400–262,144api
Gemma 3 12BNextBitint4–0.07500.1500.09380.0500–131,072api
GLM 5.3 FlashDarkbloomfp4–0.06500.1800.09380.0120–1,048,576api
Gemma 3 4BDeepInfrapriority–0.07500.1500.0938––131,072api
Schematron V2 SmallOpenRouter––0.05000.2300.09500.0500–128,000api
Schematron V2 SmallInference.net––0.05000.2300.09500.0500–128,000api
Qwen2.5 32B InstructNebius AI Studio––0.06000.2000.0950––128,000aggregator
Schematron V2 SmallInference.net–us0.05000.2300.09500.0500–128,000api
text-embedding-3-largeAzure AI Foundry––0.130free0.0975––8,191aggregator
text-embedding-3-largeOpenAI––0.130free0.0975––8,191aggregator
Reka EdgeOpenRouter––0.1000.1000.100––16,384api
Ministral 3 3B 2512OpenRouter––0.1000.1000.1000.0100–131,072api
GPT-6 Luna ProOpenAIflex–0.05000.2500.1000.0050000.06251,050,000api
GPT-6 Luna 2 sourcesOpenAIflex–0.05000.2500.1000.0050000.06251,050,000official
Reka EdgeReka AI––0.1000.1000.100––16,384api
Ministral 3 3B 2512Mistral AIzdr–0.1000.1000.1000.0100–131,072api
Ministral 3 3B 2512 3 sourcesMistral AI––0.1000.1000.1000.0100–131,072api
gpt-oss-120bCrusoe––0.05000.2500.1000.0500–131,072api
Gemma 3 27BDeepInfrafp8–0.08000.1600.100––131,072api
FW-Nemotron-Lightning-3.5-30B-A3BAzure AI Foundry––0.06000.2200.1000.0100–262,144aggregator
llama3.1-8bCerebras––0.1000.1000.100––128,000aggregator
Gemma 3 27B 3 sourcesDeepInfra––0.08000.1600.100––131,072official
eu.meta.llama3-2-1b-instruct-v1:0AWS Bedrock––0.1000.1000.100––128,000aggregator

Confidence: official = vendor pricing page · api = OpenRouter per-host · aggregator = community list, may lag · unreviewed = automatic extraction awaiting review. When sources disagree the most authoritative wins. Details: /methodology#confidence