"Unlimited Token Plan" = 50M/200M/500M tokens for $16/60/160 a month, "low priority" throttling beyond that. Includes Kimi k2.6 and MiniMax M2.5, but no OpenAI or Anthropic.
In comparison: OpenAI's 20x/$200 plan is highly subsidized. Some devs are able to get multiple billions of tokens out of it per month. Output tokens, not input or cached. Even with the most expensive model gpt-5.5 xhigh fast.
Interesting, however do you have any examples on the quality of output/results as compared to other models? Am always skeptical when I see that something is "unlimited"
"Unlimited Token Plan" = 50M/200M/500M tokens for $16/60/160 a month, "low priority" throttling beyond that. Includes Kimi k2.6 and MiniMax M2.5, but no OpenAI or Anthropic.
In comparison: OpenAI's 20x/$200 plan is highly subsidized. Some devs are able to get multiple billions of tokens out of it per month. Output tokens, not input or cached. Even with the most expensive model gpt-5.5 xhigh fast.
[dead]
Interesting, however do you have any examples on the quality of output/results as compared to other models? Am always skeptical when I see that something is "unlimited"
We'll be sharing model benchmark results in the coming weeks.