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

ModelVendorHostVariantRegionInputOutputBlended ↑?ContextConfidence?
InklingThinking MachinesOpenRouter––1.004.051.76524,288api
Inkling 3 sourcesThinking MachinesTogether AI––1.004.051.76524,288api
DeepSeek V4 ProDeepSeekAlibaba Cloudfp8–1.422.831.771,048,576api
Amazon Nova ProAmazonAWS Bedrock–eu1.054.201.84300,000aggregator
o4-mini 3 sourcesOpenAIOpenAI––1.104.401.93200,000api
o4-mini 2 sourcesOpenAIAzure AI Foundry––1.104.401.93200,000aggregator
DeepSeek V4 ProDeepSeekAlibaba Cloudfp8us-east-11.563.121.951,048,576api
DeepSeek V4 Pro 2 sourcesDeepSeekDeepSeek––1.323.961.981,048,576official
Claude Haiku 4.5AnthropicOpenRouter––1.005.002.00200,000api
GPT-6.1 Sol 2 sourcesOpenAIOpenAIflex–1.005.002.001,050,000official
DeepSeek V4 ProDeepSeekNovita AIfp8–1.603.202.001,048,576api
Claude Haiku 4.5AnthropicAzure AI Foundry–global1.005.002.00200,000api
Claude Haiku 4.5 2 sourcesAnthropicAWS Bedrock–global1.005.002.00200,000api
Claude Haiku 4.5AnthropicGoogle Vertex AI–global1.005.002.00200,000api
Claude Haiku 4.5 4 sourcesAnthropicAnthropic––1.005.002.00200,000official
Claude Haiku 4.5 2 sourcesAnthropicAzure AI Foundry––1.005.002.00200,000aggregator
Qwen3-235B-A22BAlibaba CloudHyperbolic––2.002.002.00131,072aggregator
Kimi K2Moonshot AIHyperbolic––2.002.002.00131,072aggregator
Claude Haiku 4.5AnthropicPerplexity––1.005.002.00200,000aggregator
Claude Haiku 4.5 2 sourcesAnthropicGoogle Vertex AI––1.005.002.00200,000aggregator
DeepSeek V4 Pro 2 sourcesDeepSeekNovita AI––1.603.202.001,048,576aggregator
Claude Haiku 4.5 2 sourcesAnthropicDeepInfra––1.005.002.00200,000official
GPT-6.1 SolOpenAIOpenAIbatch–1.005.002.001,050,000official
DeepSeek V4 ProDeepSeekVenice––1.653.302.061,048,576api
DeepSeek V4 ProDeepSeekAtlas Cloudfp4–1.683.382.111,048,576api
DeepSeek V4 ProDeepSeekBaidufp8–1.693.382.111,048,576api
o4-miniOpenAIAzure AI Foundry–eu1.214.842.12200,000aggregator
o4-miniOpenAIAzure AI Foundry–us1.214.842.12200,000aggregator
DeepSeek V4 ProDeepSeekStreamLake promo −45%fp8–1.743.482.181,048,576api
DeepSeek V4 ProDeepSeekGMI Cloud promo −45%fp8–1.743.482.181,048,576api
DeepSeek V4 Pro 2 sourcesDeepSeekAzure AI Foundry––1.743.482.181,048,576aggregator
DeepSeek V4 Pro 2 sourcesDeepSeekNebius AI Studio––1.753.502.191,048,576aggregator
DeepSeek V4 ProDeepSeekNebius AI Studiofp8–1.753.502.191,048,576api
Claude Haiku 4.5 2 sourcesAnthropicAWS Bedrock–us1.105.502.20200,000api
Claude Haiku 4.5AnthropicGoogle Vertex AI–us-east51.105.502.20200,000api
Claude Haiku 4.5AnthropicAWS Bedrock–eu-west-11.105.502.20200,000api
Claude Haiku 4.5AnthropicGoogle Vertex AI–europe1.105.502.20200,000api
Claude Haiku 4.5 2 sourcesAnthropicAWS Bedrock––1.105.502.20200,000aggregator
Claude Haiku 4.5AnthropicAWS Bedrock–apac1.105.502.20200,000aggregator
Claude Haiku 4.5AnthropicAWS Bedrock–au1.105.502.20200,000aggregator
Claude Haiku 4.5AnthropicAWS Bedrock–eu1.105.502.20200,000aggregator
Claude Haiku 4.5AnthropicAWS Bedrock–jp1.105.502.20200,000aggregator
DeepSeek V4 ProDeepSeekAzure AI Foundry–us1.913.832.391,048,576api
Qwen3-Max 3 sourcesAlibaba CloudDeepInfra––1.206.002.40262,144official
Claude Haiku 4.5AnthropicAWS Bedrock–us-gov-east-11.206.002.40200,000aggregator
Claude Haiku 4.5AnthropicAWS Bedrock–us-gov-west-11.206.002.40200,000aggregator
DeepSeek V4 ProDeepSeekDeepInfrafp8+priority–1.953.902.441,048,576api
Mistral Medium 3.5Mistral AIOpenRouter––1.507.503.00262,144api
Mistral Medium 3.5Mistral AIMistral AIzdr–1.507.503.00262,144api
Mistral Medium 3.5 3 sourcesMistral AIMistral AI––1.507.503.00262,144api

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