impulse

Compare models

Up to 4 models side by side: list prices, what a month of real traffic costs, limits and capabilities. The best value in each row is marked with a check.

Clear
Model
Overview
Released14 Sep 2026
Status
Current
Open weightsYes, DeepSeek License (MIT code)
Hosts
APIs you can call it on
3
Price · USD per million tokens, at the lab's own API where it sells the model
Priced atDeepSeek
Input$0.30
Output$1.20
Cached input$0.006
Batch
input / output
-
Cheapest anywhere
input / output
$0.30 / $1.20
DeepSeek
Monthly bill · at the price above, 30 days, with each workload's cache and batch use
Chat assistant
5,000 a day, 2K in, 500 out
$154
RAG over documents
2,000 a day, 8K in, 600 out
$117
Coding agent
1,000 a day, 40K in, 2K out
$185
Overnight batch
20,000 a day, 4K in, 400 out
$1,008
Limits
Context window1M tokens
Max output393K tokens
Capabilities
Takes inText, Images
ProducesText
Effort settings
as listed by the source
-
Streaming
Function calling
JSON mode
Structured outputs
Reasoning
Tool calling
Prompt caching
Assistant prefill
Parallel tool calls