Reference

Models & pricing

Cinder runs a single, fast frontier model with a very large context window.

Available models

Model IDContextStreamingToolsJSON mode
deepseek-v4-flash 1,048,576 tokens
1M context. The model accepts up to 1,048,576 tokens of combined input — long documents, large codebases, or extended conversations fit in a single call.

Pricing

Billed per token, charged against your platform balance. Prices are illustrative during preview.

ModelInput (per 1M tokens)Output (per 1M tokens)
deepseek-v4-flash$0.10$0.30

Estimating cost

A request's cost is the sum of its input and output tokens:

cost = (prompt_tokens / 1e6 * input_price)
     + (completion_tokens / 1e6 * output_price)

Every response includes a usage object with exact token counts — see the API reference. Track spend live on the platform.