Models
Every model on caft, priced at least 75% under the vendor's official list price.
Every model on caft, priced at least 75% under the vendor's official list price.
Prices are per million tokens, input / output. Open a model for the full comparison.
A monthly number of credits for a fixed price, with limits per 5 hours, per week and per month. How plans work
About 130,000 GLM-5.3-Flash, 4,500 GLM-5.3 or 710,000 DeepSeek V4.1 Flash requests a month
About 4× OpenCode Go
About 360,000 GLM-5.3-Flash, 12,000 GLM-5.3 or 1,900,000 DeepSeek V4.1 Flash requests a month
About 730,000 GLM-5.3-Flash, 24,000 GLM-5.3 or 3,900,000 DeepSeek V4.1 Flash requests a month
Plans cover GLM-5.3, GLM-5.3-Flash and DeepSeek V4.1 Flash. Other models bill your balance.
Request counts are typical coding-agent requests; actual usage varies.
Prices updated hourly. Last update at 2026-10-21 17:07:06 UTC.
Official list price is the model vendor's own published per-token price. Comparison to OpenRouter counts only providers it reports as available, excludes 4-bit (fp4) quantized offers, and assumes a ratio of 3:1 input to output tokens.