b58e377d3b1b4fff9b6ba69fce4ca1d8Q: Question about FuseBase Work limits & LLM compatibility
I am very excited for FuseBase Work but have a few questions.
1.
Is FuseBase Work included for free?
Can I set up FB Work and only pay for my API credits alongside my existing subscription limits?
I am on Tier 10 but hit a "vCPU quota limit reached" error during setup. How can I resolve this?
2.
Which LLM providers and models are supported?
The documentation mentions it currently only works with Codex and Claude Code, but also states Hermes is usable.
Which LLM API providers support Hermes on your platform?
I currently use an Open Code Go subscription (with OpenRouter as a backup) for Hermes. Can I connect these?
3.
What is the most cost-effective setup?
From a cost-to-performance perspective, is it better to use Codex, Claude Code, or Hermes (with a different API provider) within FuseBase Work?
Pavel_FUSEBASE
Jun 9, 2026A: Hi
1) Please try again. Due to high demand, we are gradually increasing the limit on the number of instances for FuseBase Work. For FuseBase Work, we provide only the instance and basic setup. For agents, you need to use your own keys or subscription.
2) Yes, you can also use Hermes, but it is currently in Beta and there may be some nuances. For Hermes, you can currently use only keys (not subscriptions) from OpenAI, Claude, Z.AI, or OpenRouter.
3) In our experience, Claude Code/Codex is sufficient for most tasks. However, this depends on the specific use case you want to address.