Our mission
Free AI models have long been underestimated. Most people dismiss them as incapable, and the companies behind them rarely invested in making free API tiers genuinely good. The best responses ended up behind paid plans, while free models were left to compete one at a time, each with its own blind spots.
Multisis exists to push free models beyond their individual limits. By combining models from many providers into a council, a Mixture of Models, we aim to deliver responses that come very close to what paid models produce, using nothing but free API keys.
We don't claim to match the best paid models in every case. Our goal is to close as much of that gap as possible, and to keep the reasoning visible so you can see where an answer came from and decide how much weight to give it.
How it works
Multisis runs every question through a council of models in two stages.
- 1
You ask a question
Type a prompt as you would in any chat. Your recent conversation is included so follow-up questions keep their context.
- 2
Three thinkers work in parallel
Three models each receive your question with a different angle to explore. They work at the same time, so you wait for the slowest one rather than the sum of all three.
- 3
A judge reconciles their work
A fourth model reads the three write-ups, resolves disagreements, removes repetition and streams a single, well-organized answer back to you.
- 4
You stay in control
Open any turn to inspect what each thinker produced, see which model handled each role and stop generation whenever you like.
Resilient by default
Principles
Transparency
You can see which model played which role and read the drafts behind every final answer.
Free models, taken further
Free models are underrated. Combined well, they can come remarkably close to paid ones, so great answers should not require a subscription.
Your keys, your choice
Multisis works with the providers you already use. You decide which models fill each role and what they cost you.
Privacy by design
Chats are private to your account, API keys are stored encrypted, and you can delete your data at any time.
Graceful failure
If a thinker or the judge is unavailable, Multisis degrades to the best available draft instead of failing outright.
Founders
Multisis is built by a small, independent team focused on making multi-model reasoning practical for everyone.
Crystal
Co-founder
I started Multisis because free models have been overlooked for too long. My vision is to show that a council of free models, drawn from many providers, can come very close to the best paid ones, so good answers are within reach of everyone, whatever they can spend.
Common questions
Why use several models instead of one?
A single model has a single set of blind spots. Asking several to approach a question from different angles, then merging the results, tends to produce answers that are more complete and better balanced.
Is it really as good as a paid model?
Not always, and we don't claim it is. A council of free models gets very close on many questions, but the strongest paid models can still pull ahead on the hardest ones. Closing that gap is what we are working on.
Does Multisis provide the AI models?
No. You connect your own provider API keys, and Multisis orchestrates them. Usage is billed by your providers according to their own terms.
Is it slower than a normal chat?
Slightly. The thinkers run in parallel, then the judge streams its answer, so total time is roughly one thinker plus one judge response.
Can I trust the final answer?
Multisis improves coverage but cannot guarantee accuracy. Always verify important information, especially for medical, legal and financial decisions.
Contact
Questions, feedback or partnership ideas are always welcome.
- General enquiries: support@multisis.space
- Read our Usage Guidelines and Privacy Policy.