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 tier-tagged and vendor-direct models. Show all offerings, including the long tail

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

ModelHostVariantRegionInputOutputBlended ↑?ContextTokenizerConfidence?
InklingOpenRouter––1.004.051.76524,288–api
Inkling 3 sourcesTogether AI––1.004.051.76524,288–api
DeepSeek V4 ProAlibaba Cloudfp8–1.422.831.771,048,576deepseekapi
Amazon Nova ProAWS Bedrock–eu1.054.201.84300,000–aggregator
o4-mini 3 sourcesOpenAI––1.104.401.93200,000o200kapi
o4-mini 2 sourcesAzure AI Foundry––1.104.401.93200,000o200kaggregator
DeepSeek V4 ProAlibaba Cloudfp8us-east-11.563.121.951,048,576deepseekapi
DeepSeek V4 Pro 2 sourcesDeepSeek––1.323.961.981,048,576deepseekofficial
Claude Haiku 4.5OpenRouter––1.005.002.00200,000claudeapi
GPT-6.1 Sol 2 sourcesOpenAIflex–1.005.002.001,050,000o200kofficial
DeepSeek V4 ProNovita AIfp8–1.603.202.001,048,576deepseekapi
Claude Haiku 4.5Azure AI Foundry–global1.005.002.00200,000claudeapi
Claude Haiku 4.5 2 sourcesAWS Bedrock–global1.005.002.00200,000claudeapi
Claude Haiku 4.5Google Vertex AI–global1.005.002.00200,000claudeapi
Claude Haiku 4.5 4 sourcesAnthropic––1.005.002.00200,000claudeofficial
Claude Haiku 4.5 2 sourcesAzure AI Foundry––1.005.002.00200,000claudeaggregator
Qwen3-235B-A22BHyperbolic––2.002.002.00131,072qwen2aggregator
Kimi K2Hyperbolic––2.002.002.00131,072–aggregator
Claude Haiku 4.5Perplexity––1.005.002.00200,000claudeaggregator
Claude Haiku 4.5 2 sourcesGoogle Vertex AI––1.005.002.00200,000claudeaggregator
DeepSeek V4 Pro 2 sourcesNovita AI––1.603.202.001,048,576deepseekaggregator
Claude Haiku 4.5 2 sourcesDeepInfra––1.005.002.00200,000claudeofficial
GPT-6.1 SolOpenAIbatch–1.005.002.001,050,000o200kofficial
DeepSeek V4 ProVenice––1.653.302.061,048,576deepseekapi
DeepSeek V4 ProAtlas Cloudfp4–1.683.382.111,048,576deepseekapi
DeepSeek V4 ProBaidufp8–1.693.382.111,048,576deepseekapi
o4-miniAzure AI Foundry–eu1.214.842.12200,000o200kaggregator
o4-miniAzure AI Foundry–us1.214.842.12200,000o200kaggregator
DeepSeek V4 ProStreamLake promo −45%fp8–1.743.482.181,048,576deepseekapi
DeepSeek V4 ProGMI Cloud promo −45%fp8–1.743.482.181,048,576deepseekapi
DeepSeek V4 Pro 2 sourcesAzure AI Foundry––1.743.482.181,048,576deepseekaggregator
DeepSeek V4 Pro 2 sourcesNebius AI Studio––1.753.502.191,048,576deepseekaggregator
DeepSeek V4 ProNebius AI Studiofp8–1.753.502.191,048,576deepseekapi
Claude Haiku 4.5 2 sourcesAWS Bedrock–us1.105.502.20200,000claudeapi
Claude Haiku 4.5Google Vertex AI–us-east51.105.502.20200,000claudeapi
Claude Haiku 4.5AWS Bedrock–eu-west-11.105.502.20200,000claudeapi
Claude Haiku 4.5Google Vertex AI–europe1.105.502.20200,000claudeapi
Claude Haiku 4.5 2 sourcesAWS Bedrock––1.105.502.20200,000claudeaggregator
Claude Haiku 4.5AWS Bedrock–apac1.105.502.20200,000claudeaggregator
Claude Haiku 4.5AWS Bedrock–au1.105.502.20200,000claudeaggregator
Claude Haiku 4.5AWS Bedrock–eu1.105.502.20200,000claudeaggregator
Claude Haiku 4.5AWS Bedrock–jp1.105.502.20200,000claudeaggregator
DeepSeek V4 ProAzure AI Foundry–us1.913.832.391,048,576deepseekapi
Qwen3-Max 3 sourcesDeepInfra––1.206.002.40262,144qwen2official
Claude Haiku 4.5AWS Bedrock–us-gov-east-11.206.002.40200,000claudeaggregator
Claude Haiku 4.5AWS Bedrock–us-gov-west-11.206.002.40200,000claudeaggregator
DeepSeek V4 ProDeepInfrafp8+priority–1.953.902.441,048,576deepseekapi
Mistral Medium 3.5OpenRouter––1.507.503.00262,144mistralapi
Mistral Medium 3.5Mistral AIzdr–1.507.503.00262,144mistralapi
Mistral Medium 3.5 3 sourcesMistral AI––1.507.503.00262,144mistralapi

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