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 ↑?Batch inBatch outContextConfidence?
GLM-4.6Zhipu AI (Z.ai)Together AI––0.6002.201.00––200,000aggregator
GLM-4.6 3 sourcesZhipu AI (Z.ai)Zhipu AI (Z.ai)––0.6002.201.00––200,000official
Llama 3.3 70B Instruct 3 sourcesMetaTogether AI––1.041.041.04––128,000api
DeepSeek V4 ProDeepSeekOpenRouter––0.9001.501.05––1,048,576api
Mistral Large 4 2 sourcesMistral AIMistral AI––0.7002.101.05––1,048,576aggregator
Kimi K2Moonshot AIOpenRouter––0.6002.501.08––131,072api
Kimi K2 3 sourcesMoonshot AINovita AI––0.6002.501.08––131,072api
Kimi K2Moonshot AINovita AIfp8–0.6002.501.08––131,072api
Kimi K2Moonshot AIAWS Bedrock––0.6002.501.080.3001.25131,072aggregator
Kimi K2Moonshot AIFireworks AI––0.6002.501.08––131,072aggregator
Kimi K2Moonshot AIAWS Bedrock–us-east-10.6002.501.08––131,072aggregator
Kimi K2Moonshot AIAWS Bedrock–us-east-20.6002.501.08––131,072aggregator
Kimi K2Moonshot AIAWS Bedrock–us-west-20.6002.501.08––131,072aggregator
Kimi K2Moonshot AIAWS Bedrock–ap-southeast-20.6182.581.11––131,072aggregator
DeepSeek V4 ProDeepSeekSiliconFlowfp8–1.001.501.13––1,048,576api
DeepSeek V3.2 (reasoner)DeepSeekDeepInfra––0.7002.401.13––128,000aggregator
DeepSeek V3.2 (reasoner)DeepSeekNovita AIfp8–0.7002.501.15––128,000api
DeepSeek V3.2 (reasoner)DeepSeekNebius AI Studio––0.8002.401.20––128,000aggregator
DeepSeek V4 ProDeepSeekFireworks AI––1.201.201.20––1,048,576aggregator
Qwen3-235B-A22BAlibaba CloudOpenRouter––0.4503.501.21––131,072api
Qwen3-235B-A22BAlibaba CloudVenicefp8–0.4503.501.21––131,072api
Qwen3-Coder-480B-A35BAlibaba CloudAWS Bedrock–eu-west-20.7002.791.22––262,144aggregator
Kimi K2Moonshot AIAWS Bedrock–ap-south-10.7102.941.27––131,072aggregator
Kimi K2Moonshot AIAWS Bedrock–ap-southeast-30.7203.001.29––131,072aggregator
Kimi K2Moonshot AIAWS Bedrock–eu-north-10.7203.001.29––131,072aggregator
DeepSeek V4 ProDeepSeekDeepInfraflex+fp8–1.042.081.30––1,048,576api
DeepSeek V4 ProDeepSeekDigitalOcean––1.042.091.31––1,048,576api
Kimi K2Moonshot AIAWS Bedrock–ap-northeast-10.7303.031.31––131,072aggregator
Kimi K2Moonshot AIAWS Bedrock–sa-east-10.7303.031.31––131,072aggregator
Amazon Nova Pro 2 sourcesAmazonAWS Bedrock––0.8003.201.400.4001.60300,000aggregator
Amazon Nova ProAmazonAmazon––0.8003.201.40––300,000aggregator
Amazon Nova ProAmazonAWS Bedrock–us0.8003.201.400.4001.60300,000aggregator
Step 5 PreviewStepFunStepFunfp8–1.002.701.43––1,000,000api
Step 5 PreviewStepFunOpenRouter––1.002.701.43––1,000,000api
Step 5 PreviewStepFunNovita AI––1.002.701.43––1,000,000aggregator
Step 5 PreviewStepFunfeatherlessfp4–1.002.701.43––1,000,000api
Step 5 PreviewStepFunDeepInfra––1.002.701.43––1,000,000aggregator
Step 5 PreviewStepFunGMI Cloudfp8–1.002.701.43––1,000,000api
Amazon Nova ProAmazonAWS Bedrock–apac0.8403.361.47––300,000aggregator
Kimi K2Moonshot AITogether AI––1.003.001.50––131,072aggregator
Gemini 2.5 FlashGoogleGooglepriority–0.5404.501.53––1,048,576api
Qwen3-Max 3 sourcesAlibaba CloudAlibaba Cloud––0.7803.901.56––262,144api
DeepSeek V4 ProDeepSeekDeepInfrafp8–1.302.601.63––1,048,576api
DeepSeek V4 Pro 2 sourcesDeepSeekDeepInfra––1.302.601.63––1,048,576aggregator
Kimi K2Moonshot AIAWS Bedrock–eu-west-20.9303.881.67––131,072aggregator
Amazon Nova ProAmazonAWS Bedrock–us-gov-east-10.9603.841.68––300,000aggregator
Amazon Nova ProAmazonAWS Bedrock–us-gov-west-10.9603.841.680.4801.92300,000aggregator
GPT-5.1OpenAIOpenAIflex–0.6255.001.72––400,000api
Gemini 2.5 ProGoogleGoogleflex–0.6255.001.72––1,048,576api
Inkling 2 sourcesThinking MachinesDeepInfra––0.9504.051.73––524,288aggregator

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