Basis is live · custom pricing · last verified 9 Nov 2026 · viability monitor
Basis
Basis is a nonprofit research lab building multi-agent coordination and machine-checked formal verification, not a product you buy.
Fund or collaborate with Basis if open research into agent coordination and formal verification matters to your roadmap. The nftables bug patches and the October 2026 Lean-on-RISC-V interpreter proof are checkable output rather than slideware. If you need an API to call next sprint, pass — this is a research bet priced in researcher time, not licenses.
Verdict generatedVerified 5d ago · liveness 68/100 · cite: rightaichoice.com/tools/basis
- AI researchers working on multi-agent coordination, formal verification or agent trust
- Academic labs seeking an open-science research partner with shared authorship
- Philanthropies funding public-interest AI research with checkable output
- Research fellows who want to join an engineering-heavy, proof-driven lab
- Developers who need a production API or SDK to call today
- Businesses needing a packaged product with support terms or an SLA
- Teams that require procurement contracts or uptime guarantees
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Basis if you need a product you can contract, integrate and put into production this quarter — it is a nonprofit lab publishing research, benchmarks and open-source methods, not a vendor with a delivered API.
There is no product to price. Budget for Basis looks like a grant, donation or sponsored research agreement, which puts it in a different category from paid AI tooling: cheaper than a commercial platform contract if you are funding a small piece of open research, far more expensive in researcher time if what you actually needed was software you could deploy.
In short
Basis — Basis is a nonprofit research lab building multi-agent coordination and machine-checked formal verification, not a product you buy. Best for AI researchers working on multi-agent coordination, formal verification or agent trust, Academic labs seeking an open-science research partner with shared authorship, Philanthropies funding public-interest AI research with checkable output. Pricing is on request (custom quotes).
No head-to-head comparison for Basis yet — the alternatives below are the closest matches we list.
What is Basis?
Basis — Basis is a nonprofit research lab building multi-agent coordination and machine-checked formal verification, not a product you buy. Best for AI researchers working on multi-agent coordination, formal verification or agent trust, Academic labs seeking an open-science research partner with shared authorship, Philanthropies funding public-interest AI research with checkable output. Pricing is on request (custom quotes).
Basis is a nonprofit applied AI research lab, not a software vendor. Its site is organised around Core Technology, Challenge Projects, Modules, Fellowships, Collaborators and Offerings, and the lab itself describes its mission as creating "a new kind of AI research organization." The ambition is a universal reasoning engine — framed as a research goal, not a product you can license. Two named threads anchor the Core Technology work. Pact is a formal coordination language that fuses game theory, distributed systems and cryptography so autonomous agents can be trusted to cooperate. ExoPredicator learns abstract world models that let robots plan in time-varying environments by abstracting over time and state. Around those sit evaluation artifacts: AutumnBench, a benchmark with human baselines for agent behaviour, and the MARA protocol for testing causal reasoning in agents. Recent output is specific and checkable. In October 2026 the lab published work in which LLMs wrote formal semantics for a small language, with Lean proving the compiled interpreter matches them on RISC-V; a companion post shipped tools for reconstructing animal movement in 3D from field video. An earlier project turned LLM-based verification on the Linux nftables nft CLI and patched kernel bugs that had survived years of review, and another had four agents build a JavaScript-to-WASM compiler in Lean over 14 days. The audience is narrow: research engineers, academics, and funders who value open, machine-checked results. Judged against commercial agent-coordination vendors, Basis is a place to read methods, cite benchmarks, and possibly collaborate — not a service with an API contract or an SLA.
Key features
- Pact: formal coordination language for multi-agent systems
- Pact blends game theory, distributed systems and cryptography for agent trust
- ExoPredicator: learned abstract world models for robot planning
- ExoPredicator abstracts time and state to plan in dynamic environments
- LLM-based verification of production system software (Linux nftables nft CLI)
- LLM-written formal semantics for a small language, proved in Lean on RISC-V
- Four agents built a JavaScript-to-WASM compiler in Lean over 14 days
- AutumnBench benchmark with human baselines for agent behaviour
- MARA protocol for testing causal reasoning in agents
- Tools for reconstructing animal movement in 3D from field video
- Challenge-driven research projects organised as modules
- Open publication of papers and research findings
- Open-source code releases
- Fellowship program for researchers
- Partnership and collaboration programs
Behind the verdict
We'd point research engineers and funders at Basis when the question is "how do we make autonomous agents trustworthy enough to coordinate," and they want methods with proofs attached rather than a vendor deck. The nftables work is the convincing artefact: LLM-based verification found and patched kernel bugs that human review had missed for years. That is a result you can audit. Pick Basis if you want to cite a benchmark (AutumnBench, MARA), collaborate on formal methods, or fund public-interest AI with checkable output. The October 2026 Lean proofs that a compiled interpreter matches LLM-written semantics on RISC-V, and the animal-tracking tools from field video, show the lab ships real code alongside the theory. Pass if you need a signed API contract this quarter, an SDK, or uptime guarantees. There is nothing here to buy. Compare against commercial agent-coordination vendors for a packaged service; come to Basis for the underlying Pact language and the research record. In practice, budget researcher time rather than licences, and expect to read papers and repos rather than plug into a dashboard. The payoff is proof-driven work you can build on; the cost is that timelines follow research, not a product roadmap.
Limitations
- Basis is a 501(c)(3) nonprofit applied research organisation.
- Its published outputs are research papers, benchmarks, protocols and open-source software rather than a procurable end-user product.
- Funding is framed around grants, donations and company contracts, and the public contact route is a general address (contact@basis.ai).
- Named work spans formal methods and verification (the Pact coordination language, LLM-based verification of Linux's nftables nft CLI, an LLM-written semantics and Lean interpreter proof on RISC-V), robot planning (ExoPredicator), animal-movement reconstruction from field video, and benchmarks such as AutumnBench and the MARA protocol.
- Treat Basis as a research collaborator rather than a vendor.
as of 2026-10-25
Researching Basis? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Is Basis still active?
Basis is live · custom pricing · last verified 9 Nov 2026 · viability monitor.
Viability score
How well maintained and how widely used is Basis? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: November 2026
How we score →What's new in Basis
Checked 8 days agoAcross the latest 2 updates: 2 news mentions.
Basis: LLMs wrote formal semantics for a small language, proved with Lean on RISC-V
Basis says LLMs wrote formal semantics for a small programming language and Lean proofs that the interpreter follows them after compilation to RISC-V.
Basis technical post: reconstructing animal movements in 3D from field video
Basis published practical tools for reconstructing animal tracks and their surroundings from outdoor video footage.
What do users say about Basis?
What people actually say about Basis — is it worth it?
We scanned public community sources for Basis on Oct 11, 2026 and could not establish that the discussion we found is about this tool rather than something else sharing its name. Only 2 of the posts we fetched could be positively tied to Basis. Rather than publish a sentiment score built on the wrong subject, we publish nothing here and re-run the scan.
How we verified this
Every listed fact about Basis — pricing, features, links, status — is re-read from the vendor's own pages by our verification pipeline and each change is logged with a date; the last pass ran on 9 Nov 2026. Analysis is drafted with AI models and checked against our editorial standards before publishing; see how we verify.
Verification history
We have re-verified Basis 24 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — changed description, our verdict, summary
- — changed pricing page
- — changed who it suits, summary, who should skip it, our verdict, description, features
- — changed SSL certificate
- — changed use cases, who should skip it, switching in, switching out, description, who it suits, summary, our verdict, setup time, pricing fit, features, hidden costs, limitations
- — changed limitations, pricing page, features, who should skip it, summary, description, who it suits, our verdict
Showing the 6 most recent of 24 verification passes.
Free to cite with attribution — this page re-verifies continuously.
How much does Basis cost in 2026?
Basis: Pricing is on request (custom quotes). Pricing last verified 9 Nov 2026.
Where the pricing makes sense
The company stage and team size where Basis's pricing actually pencils out — and where peers do it cheaper.
There is no product to price. Budget for Basis looks like a grant, donation or sponsored research agreement, which puts it in a different category from paid AI tooling: cheaper than a commercial platform contract if you are funding a small piece of open research, far more expensive in researcher time if what you actually needed was software you could deploy.
Has Basis' pricing changed?
No plan price for Basis has moved since we began logging in Aug 2026; the pricing page is re-read on every verification pass.
Who should pick Basis?
Pick Basis if you are AI researchers working on multi-agent coordination, formal verification or agent trust, Academic labs seeking an open-science research partner with shared authorship, Philanthropies funding public-interest AI research with checkable output, Research fellows who want to join an engineering-heavy, proof-driven lab.
Look elsewhere if you are Developers who need a production API or SDK to call today, Businesses needing a packaged product with support terms or an SLA, Teams that require procurement contracts or uptime guarantees, Consumers looking for a chatbot or a consumer app.
Skip it if Skip Basis if you need a product you can contract, integrate and put into production this quarter — it is a nonprofit lab publishing research, benchmarks and open-source methods, not a vendor with a delivered API.
GeneratedReal-world workflow fit
Concrete scenarios for the personas Basis actually fits — and what changes day-one when you adopt it.
You want a formal basis for agent trust rather than ad-hoc prompt scaffolding, so you read the Pact work and run your coordination experiments against it, then compare results with the AutumnBench human baselines.
Outcome: You get a citable formal framework plus a baseline to measure against, and a route to approaching the lab as a collaborator rather than a customer.
You are looking for public-interest AI work with checkable output, so you review the named results — the nftables kernel bug patches, the Lean interpreter proof on RISC-V — and the module structure before starting a conversation about support.
Outcome: You fund deliverables you can verify rather than roadmaps, with papers and open-source releases as the receipt.
You want to see how far LLM-assisted verification reaches into production software, so you follow the nftables work and the agent-built JavaScript-to-WASM compiler in Lean to inform your own tooling choices.
Outcome: You get worked examples of LLM-plus-proof pipelines applied to real systems code, without waiting for a vendor to productise them.
Use cases
- Research multi-agent coordination and coordination languages using Pact
- Plan robot behaviour with learned abstract world models via ExoPredicator
- Benchmark world-model learning algorithms against human baselines with AutumnBench
- Test causal reasoning in agents with the MARA protocol
- Apply LLM-based verification to real system software, as with Linux's nftables nft CLI
- Study program synthesis and formal verification, including agent-built compilers in Lean
- Reconstruct animal movement and surroundings in 3D from outdoor field video
- Collaborate as a research fellow on mission-driven open-science projects
Setup time & first value
How long it actually takes to get something useful out of Basis — broken out by persona, not the marketing-page minute.
For a reader: minutes — the site is thin navigation to work that mostly lives in papers and code. For a collaborator: weeks of reading before a useful conversation. For a fellow: a formal application cycle. Nothing here is installed or configured.
Resources & guides
Tutorials & Learning
YouTube returned 6 videos for “Basis”, and we withheld 6: 6 could not be judged, because “Basis” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Basis.
Official links
What are the best Basis alternatives?
The closest alternatives to Basis we list are Spine AI, ScienceClaw and GPT Researcher. Ranked by direct product-type match, not category overlap.
| Alternative | Pricing | Starts at | Last verified |
|---|---|---|---|
| Spine AI | Freemium | Pricing not published | — |
| ScienceClaw | Free | Free | — |
| GPT Researcher | Freemium | Pricing not published | — |
Alternatives to Basis
View all
Spine AI
Multi-agent research that turns a question into a sourced deck, Excel model, or app
ScienceClaw
ScienceClaw is a loop-engineered research agent that plans, executes, checks and iterates on multi-step scientific tasks.
GPT Researcher
Open-source deep research agent that reads sources in parallel and writes cited reports you self-host
Tools that pair well with Basis
Common stack mates teams adopt alongside Basis, with the specific reason each pairing earns its keep.
Spine AI
Multi-agent research that turns a question into a sourced deck, Excel model, or app
ScienceClaw
ScienceClaw is a loop-engineered research agent that plans, executes, checks and iterates on multi-step scientific tasks.
GPT Researcher
Open-source deep research agent that reads sources in parallel and writes cited reports you self-host
Frequently asked questions
Categories
Best-of guides
For makers of Basis
- Get a Verified Listing for Basis ($39): a written decision within 1 working day.
- Free listing for another tool: Free reviews are decided in order, within 3 weeks of the review starting.
Used Basis? Help shape our editorial sentiment research.