Undocumented 100 RPM limit on Go plan — rate limits for higher tiers?

[984babylon]

Hi OpenDesign team,

I'm currently subscribed to the Go plan and using DeepSeek V4.1 Flash through OpenCode.

When running five subagents concurrently, all of them failed due to the following error:

400: {
  "code": "tier_limit_exceeded",
  "message": "membership rpm limit exceeded: 101/100 resets 2026-10-19T23:18:00Z",
  "type": "tier_limit_exceeded"
}

This indicates a limit of 100 requests per minute (RPM) for my account.

I noticed that the pricing page advertises up to 10 concurrent tasks, but I couldn't find any documentation specifying this RPM limit.

Could you clarify the following?

1. Is 100 RPM the standard limit for all Go plan subscribers?
2. What are the RPM limits for the higher-tier plans (Plus and Pro)?
3. Are there additional limits, such as tokens per minute (TPM) or maximum concurrent API requests?
4. Is there any way to increase the RPM limit?
5. How is the advertised support for 10 concurrent tasks intended to work with the 100 RPM limit?
6. Could these limits be documented on the pricing page so users can evaluate the plans before subscribing?

I understand that rate limits are necessary to maintain service stability. However, for users running multiple coding agents, RPM limits can be more important than the advertised monthly token allowance.

I'd appreciate clarification on the limits for each plan.

Thanks!