FutureSearch
Summary
Most forecasting tools hand you a probability and leave you to wonder whether it came from a rigorous model or a vibes-calibrated heuristic — FutureSearch publishes its full track record, including losing positions, so you can audit the accuracy before you bet a business decision on it.
FutureSearch runs multi-agent research teams behind each query, returning probability, numeric, date, categorical, conditional, and decision forecasts. The vendor states a #1 ranking of 226 on the Metaculus FutureEval live tournament and a #20 of 335 placement on the Forecasting Research Institute's ForecastBench — a contamination-free benchmark, which matters when other tools train on the answers. Per-question pricing scales with effort, so a quick probability check costs less than a deep conditional scenario. The tool is SaaS-only with no self-hosted option, which is a hard stop for teams with data residency requirements. API access and an MCP connector for Claude are documented, making integration into existing AI workflows direct.
Bottom line: Reach for FutureSearch when you need a defensible, benchmark-verified probability on a market or business question you can act on — but if your organization cannot send question content to an external SaaS endpoint, the architecture is a non-starter.
Pricing Plans
Usage-Based- Price
- $0.15 to $2 per question
- Free Tier
- $20 free credits on signup
Research plan
Half price for first month
View full pricing on futuresearch.ai →
Pricing may have changed since last verified. Check the official site for current plans.
Community Performance Report Card
No community ratings yet. Be the first to rate this tool!
Pros
Sign in to edit- Third-party benchmark placement on ForecastBench and Metaculus EvalTournament, so you are comparing against a contamination-free public leaderboard rather than vendor-selected demos — which means you have evidence to cite when a stakeholder asks why you trust the number.
- Six forecast types (probability, numeric, date, categorical, conditional, decision) in a single endpoint, so a team modeling an IPO timeline and its conditional effect on ARR does not stitch together separate tools.
- Multi-agent research teams run behind each query, which means the output includes the reasoning chain — not just the probability — so you can audit whether the model read the right sources before you act on the estimate.
- API and MCP connector support for Claude Code, Cursor, and ChatGPT Codex, so forecasting becomes a callable function inside pipelines that already exist rather than a tab you switch to manually.
- Live trading positions published with the research behind them, including losses, which means the stated accuracy is checkable against real-money outcomes rather than backtested cherry-picks.
Cons
Sign in to edit- No self-hosted deployment option exists. Teams with data residency obligations or policies against sending proprietary business questions to external SaaS endpoints cannot use the tool at all — the architecture provides no path around this, and those teams move to constrained local forecasting pipelines or domain-specific models they can host themselves.
- Per-question pricing scales with effort, and high-effort conditional or decision questions sit at the top of the cost range. Teams running forecasting at volume — hundreds of questions per week as a screening layer — will need to model usage costs carefully; the free credits cover exploration, not production throughput.
- Benchmark scores place FutureSearch above frontier model baselines on the BTF-3 pastcasting task, but the best frontier model scores 0.130 versus FutureSearch's 0.122, meaning for teams who already have a frontier model embedded in their stack, the marginal accuracy gain is measurable but not an order-of-magnitude jump — the business case rests on calibration and reasoning transparency, not raw score dominance.
About
- API Available
- Yes
- Self-Hosted
- No
- Last Updated
- 2026-08-16T11:52:47.838Z
Best For
Who it's for
- Users needing probability and numeric forecasts
- Teams requiring documented benchmark performance
- Organizations integrating forecasting into existing AI workflows
What it does well
- Forecasting annual records and market events
- Estimating company financial metrics such as ARR
- Predicting timing of corporate actions like IPOs
- Evaluating conditional scenarios such as geopolitical impacts on prices
- Assessing internal business decisions on growth targets
Integrations
Add notes, reviews, and benchmarks so the next visitor gets a clearer picture.
Compare FutureSearch
Spotted incorrect or missing data? Join our community of contributors.
Sign Up to ContributeRecommended skills for this tool
Auto-curated by the AIDiveForge recommendation matrix. These skills are predicted to enhance this tool based on category, capability, and domain signals.
-
Meeting Summary Template transform 32%
Turn a raw transcript into a decision-focused recap: outcomes, owners, deadlines, open threads.
Why: category partial · caps 0/0 · domain ops
-
Standup Note Synthesizer transform 32%
Merge individual standup bullets from multiple people into a single team digest with blockers surfaced to the top.
Why: category partial · caps 0/0 · domain ops
-
Runbook Skeleton post 32%
Produce a first-draft runbook from a postmortem — detection, diagnosis, mitigation, rollback — so the next incident has a template to follow.
Why: category partial · caps 0/0 · domain ops
-
OKR Draft Critiquer post 32%
Score draft OKRs against SMART criteria and the outcome-not-output rule, with suggested rewrites for each failing key result.
Why: category partial · caps 0/0 · domain ops
Frequently Asked Questions
- Is FutureSearch free?
- FutureSearch has a permanent free tier alongside paid upgrades (paid plans from $0.15 to $2 per question). You can keep using a baseline version indefinitely without paying.
- Is FutureSearch open source?
- No — FutureSearch is a closed-source tool. Source code is not publicly available.
- Does FutureSearch have an API?
- Yes. FutureSearch exposes a developer API. See the official documentation at https://futuresearch.ai for details.
Best FutureSearch alternatives →
Curated lists that include this category
Most forecasting tools hand you a probability without showing how it was made
FutureSearch runs multi-agent research teams behind each query and returns probability, numeric, date, categorical, conditional, and decision forecasts. The vendor states a #1 ranking of 226 on the Metaculus FutureEval live tournament and a #20 of 335 placement on the Forecasting Research Institute’s ForecastBench. Pricing runs $0.15 to $2 per question on a usage-based model, with $20 free credits on signup. The tool supports six forecast types through a single endpoint and integrates with Claude, Cursor, ChatGPT, Codex, Claude.ai, and an MCP connector. It is SaaS-only with no self-hosted option.
Who it is for / who should skip it
Best for users needing probability and numeric forecasts or teams that want documented benchmark performance they can cite. Organizations already working inside Claude or ChatGPT workflows can add it without new infrastructure. Skip it if your team has data-residency rules that block external SaaS or if you expect to run hundreds of high-effort conditional questions per week, since costs scale with effort and the architecture offers no local path.