Forward Future Tools Library
MiroFish
MiroFish turns source material and a decision question into multi-agent scenario simulations and inspectable reports for teams rehearsing stakeholder, market, policy, or narrative reactions.
Try MiroFish →
mirofish.work·Free + paid plans·Checked 2026-09-29
›What is MiroFish?
MiroFish is an AI prediction engine that uses knowledge graphs and multi-agent simulation to explore what might happen in a defined scenario. Users can upload text, PDFs, Markdown, TXT files, or decision notes, then review predicted reactions, branch points, timelines, probabilities, and follow-up paths. Reports can be questioned through natural-language chat with simulated agents or a ReportAgent.
›What are the pros and cons of MiroFish?
Strengths
Turns a prediction question into a staged simulation rather than a one-shot answer
Grounds simulations in uploaded source material and extracted entity relationships
Provides inspectable reports with branch points, risk signals, timelines, and follow-up paths
Lets users question simulated agents after the initial run
Supports variable injection and scenario reruns when assumptions change
Trade-offs
Free accounts are limited to one guest simulation or two signed-in simulations and DeepSeek V4 Flash
Starter does not include a knowledge base, custom-material simulations, or API access
Paid plans cap monthly analyses and the number of agents per simulation
Public materials do not establish forecast accuracy for high-stakes decisions
Self-hosted deployment requires multiple services and credentials
›What are MiroFish’s key features?
Builds a GraphRAG knowledge graph from uploaded source material
Runs independent AI agents in a simulated social environment
Accepts plain-language scenario questions and handles setup, simulation, and reporting in one conversation
Produces structured reports with group dynamics, probabilities, timeline notes, and follow-up paths
Injects policy changes, market shocks, or other variables to rerun a scenario
Supports post-simulation questions to agents and the ReportAgent
›What are the best use cases for MiroFish?
Rehearse how customers may respond to a product price change
Explore public reactions to a brand spokesperson change or policy debate
Compare market, campaign, or crisis-response scenarios
Analyze how narratives may spread among different stakeholder groups
Simulate story-world and character outcomes
›What is the pricing for MiroFish?
| Plan | Price | Details |
|---|---|---|
| Free | $0/trial | One guest simulation or two simulations for signed-in free accounts, using DeepSeek V4 Flash only. |
| Starter | $9/month | 10 analyses per month and up to 50 agents per simulation. |
| Pro | $29/month | 50 analyses per month, up to 100 agents per simulation, knowledge base, custom-material simulation, and API access. |
| Enterprise | $59/month | 320 analyses per month, up to 200 agents per simulation, knowledge base, custom-material simulation, and API access. |
Yearly billing is $54/year for Starter, $174/year for Pro, and $354/year for Enterprise. Free guest and signed-in allowances differ.
Checked 2026-09-29 · source
›Who is MiroFish best for?
analystsUseful for comparing stakeholder, market, and policy scenarios through repeatable simulation reports.
small teamPro supports repeated scenario comparisons with custom materials, a knowledge base, and API access.
enterpriseEnterprise provides higher monthly analysis capacity and simulations of up to 200 agents.
developersAPI access and inspectable agent interactions support teams integrating scenario runs into technical workflows.
Not for
- Teams that need proven forecast accuracy for high-stakes decisions should validate outputs with outside evidence.
- Buyers seeking measured human evidence should use surveys, interviews, analytics, or experiments instead.
- Teams that only need quick drafting or explanation may find a standard chatbot simpler.
›What are the best MiroFish alternatives?
AI Predictor EngineChoose this instead if you need explicit optimistic, base, and bearish scenario paths with probability and confidence outputs.→
ScenarioPlanning AIChoose this instead if your main work is planning product launches, pricing changes, and competitive moves in a dedicated scenario workspace.→
Sema4.ai StudioChoose this instead if you need to create, test, and deploy enterprise AI agents using runbooks rather than simulate stakeholder reactions.→
AionUi Cowork Console→