Verified LLM API prices per million tokens
InferenceRightNow.com compares dated input and output token prices for hosted language models. Current model coverage includes gpt-oss-120b, Llama 3.3 70B Instruct, DeepSeek V3.2, DeepSeek R1-0528, Qwen3-235B-A22B Instruct 2507, Kimi K2.6, and Kimi K3.
Prices are shown in USD per 1 million input and output tokens, with provider source links and a verification date. An estimator calculates monthly cost from the visitor’s token mix and highlights the lowest verified price for each model. Providers without a verified public number are labeled “check live.”
Dedicated inference
Teams spending $3,000 or more per month on tokens, or needing a dedicated endpoint, can request quotes. The request form prepares an email draft; the visitor reviews and sends it. Quotes are coordinated with inference providers, and provider terms determine final pricing.
Pricing methodology
Prices are manually checked against provider pricing pages on the date shown. Dedicated, batch, priority, and cached-input rates are excluded unless labeled. Provider links and live-price links lead to official provider sites.
Contact
Dedicated-inference quote requests: quotes@inferencerightnow.com.
Model price comparison · How prices are checked · Request dedicated-inference quotes · Disclosure · GPU Right Now for GPU-hour prices and deployment.