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
| Model | Host | Variant | Region | Input | Output | Blended ↑? | Batch in | Batch out | Context | Tokenizer | Confidence? |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Inkling | OpenRouter | – | – | 1.00 | 4.05 | 1.76 | – | – | 524,288 | – | api |
| Inkling 3 sources | Together AI | – | – | 1.00 | 4.05 | 1.76 | – | – | 524,288 | – | api |
| DeepSeek V4 Pro | Alibaba Cloud | fp8 | – | 1.42 | 2.83 | 1.77 | – | – | 1,048,576 | deepseek | api |
| Amazon Nova Pro | AWS Bedrock | – | eu | 1.05 | 4.20 | 1.84 | – | – | 300,000 | – | aggregator |
| o4-mini 3 sources | OpenAI | – | – | 1.10 | 4.40 | 1.93 | 0.550 | 2.20 | 200,000 | o200k | api |
| o4-mini 2 sources | Azure AI Foundry | – | – | 1.10 | 4.40 | 1.93 | 0.550 | 2.20 | 200,000 | o200k | aggregator |
| DeepSeek V4 Pro | Alibaba Cloud | fp8 | us-east-1 | 1.56 | 3.12 | 1.95 | – | – | 1,048,576 | deepseek | api |
| DeepSeek V4 Pro 2 sources | DeepSeek | – | – | 1.32 | 3.96 | 1.98 | – | – | 1,048,576 | deepseek | official |
| Claude Haiku 4.5 | OpenRouter | – | – | 1.00 | 5.00 | 2.00 | – | – | 200,000 | claude | api |
| GPT-6.1 Sol 2 sources | OpenAI | flex | – | 1.00 | 5.00 | 2.00 | – | – | 1,050,000 | o200k | official |
| DeepSeek V4 Pro | Novita AI | fp8 | – | 1.60 | 3.20 | 2.00 | – | – | 1,048,576 | deepseek | api |
| Claude Haiku 4.5 | Azure AI Foundry | – | global | 1.00 | 5.00 | 2.00 | – | – | 200,000 | claude | api |
| Claude Haiku 4.5 2 sources | AWS Bedrock | – | global | 1.00 | 5.00 | 2.00 | 0.500 | 2.50 | 200,000 | claude | api |
| Claude Haiku 4.5 | Google Vertex AI | – | global | 1.00 | 5.00 | 2.00 | – | – | 200,000 | claude | api |
| Claude Haiku 4.5 4 sources | Anthropic | – | – | 1.00 | 5.00 | 2.00 | 0.500 | 2.50 | 200,000 | claude | official |
| Claude Haiku 4.5 2 sources | Azure AI Foundry | – | – | 1.00 | 5.00 | 2.00 | – | – | 200,000 | claude | aggregator |
| Qwen3-235B-A22B | Hyperbolic | – | – | 2.00 | 2.00 | 2.00 | – | – | 131,072 | qwen2 | aggregator |
| Kimi K2 | Hyperbolic | – | – | 2.00 | 2.00 | 2.00 | – | – | 131,072 | – | aggregator |
| Claude Haiku 4.5 | Perplexity | – | – | 1.00 | 5.00 | 2.00 | – | – | 200,000 | claude | aggregator |
| Claude Haiku 4.5 2 sources | Google Vertex AI | – | – | 1.00 | 5.00 | 2.00 | 0.500 | 2.50 | 200,000 | claude | aggregator |
| DeepSeek V4 Pro 2 sources | Novita AI | – | – | 1.60 | 3.20 | 2.00 | – | – | 1,048,576 | deepseek | aggregator |
| Claude Haiku 4.5 2 sources | DeepInfra | – | – | 1.00 | 5.00 | 2.00 | – | – | 200,000 | claude | official |
| GPT-6.1 Sol | OpenAI | batch | – | 1.00 | 5.00 | 2.00 | – | – | 1,050,000 | o200k | official |
| DeepSeek V4 Pro | Venice | – | – | 1.65 | 3.30 | 2.06 | – | – | 1,048,576 | deepseek | api |
| DeepSeek V4 Pro | Atlas Cloud | fp4 | – | 1.68 | 3.38 | 2.11 | – | – | 1,048,576 | deepseek | api |
| DeepSeek V4 Pro | Baidu | fp8 | – | 1.69 | 3.38 | 2.11 | – | – | 1,048,576 | deepseek | api |
| o4-mini | Azure AI Foundry | – | eu | 1.21 | 4.84 | 2.12 | 0.605 | 2.42 | 200,000 | o200k | aggregator |
| o4-mini | Azure AI Foundry | – | us | 1.21 | 4.84 | 2.12 | 0.605 | 2.42 | 200,000 | o200k | aggregator |
| DeepSeek V4 Pro | StreamLake promo −45% | fp8 | – | 1.74 | 3.48 | 2.18 | – | – | 1,048,576 | deepseek | api |
| DeepSeek V4 Pro | GMI Cloud promo −45% | fp8 | – | 1.74 | 3.48 | 2.18 | – | – | 1,048,576 | deepseek | api |
| DeepSeek V4 Pro 2 sources | Azure AI Foundry | – | – | 1.74 | 3.48 | 2.18 | – | – | 1,048,576 | deepseek | aggregator |
| DeepSeek V4 Pro 2 sources | Nebius AI Studio | – | – | 1.75 | 3.50 | 2.19 | – | – | 1,048,576 | deepseek | aggregator |
| DeepSeek V4 Pro | Nebius AI Studio | fp8 | – | 1.75 | 3.50 | 2.19 | – | – | 1,048,576 | deepseek | api |
| Claude Haiku 4.5 2 sources | AWS Bedrock | – | us | 1.10 | 5.50 | 2.20 | 0.550 | 2.75 | 200,000 | claude | api |
| Claude Haiku 4.5 | Google Vertex AI | – | us-east5 | 1.10 | 5.50 | 2.20 | – | – | 200,000 | claude | api |
| Claude Haiku 4.5 | AWS Bedrock | – | eu-west-1 | 1.10 | 5.50 | 2.20 | – | – | 200,000 | claude | api |
| Claude Haiku 4.5 | Google Vertex AI | – | europe | 1.10 | 5.50 | 2.20 | – | – | 200,000 | claude | api |
| Claude Haiku 4.5 2 sources | AWS Bedrock | – | – | 1.10 | 5.50 | 2.20 | 0.550 | 2.75 | 200,000 | claude | aggregator |
| Claude Haiku 4.5 | AWS Bedrock | – | apac | 1.10 | 5.50 | 2.20 | 0.550 | 2.75 | 200,000 | claude | aggregator |
| Claude Haiku 4.5 | AWS Bedrock | – | au | 1.10 | 5.50 | 2.20 | 0.550 | 2.75 | 200,000 | claude | aggregator |
| Claude Haiku 4.5 | AWS Bedrock | – | eu | 1.10 | 5.50 | 2.20 | 0.550 | 2.75 | 200,000 | claude | aggregator |
| Claude Haiku 4.5 | AWS Bedrock | – | jp | 1.10 | 5.50 | 2.20 | 0.550 | 2.75 | 200,000 | claude | aggregator |
| DeepSeek V4 Pro | Azure AI Foundry | – | us | 1.91 | 3.83 | 2.39 | – | – | 1,048,576 | deepseek | api |
| Qwen3-Max 3 sources | DeepInfra | – | – | 1.20 | 6.00 | 2.40 | – | – | 262,144 | qwen2 | official |
| Claude Haiku 4.5 | AWS Bedrock | – | us-gov-east-1 | 1.20 | 6.00 | 2.40 | – | – | 200,000 | claude | aggregator |
| Claude Haiku 4.5 | AWS Bedrock | – | us-gov-west-1 | 1.20 | 6.00 | 2.40 | – | – | 200,000 | claude | aggregator |
| DeepSeek V4 Pro | DeepInfra | fp8+priority | – | 1.95 | 3.90 | 2.44 | – | – | 1,048,576 | deepseek | api |
| Mistral Medium 3.5 | OpenRouter | – | – | 1.50 | 7.50 | 3.00 | – | – | 262,144 | mistral | api |
| Mistral Medium 3.5 | Mistral AI | zdr | – | 1.50 | 7.50 | 3.00 | – | – | 262,144 | mistral | api |
| Mistral Medium 3.5 3 sources | Mistral AI | – | – | 1.50 | 7.50 | 3.00 | – | – | 262,144 | mistral | api |
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