--- title: "Feynman AI Research Agent Review: Costs & Fit" canonical: https://wavect.io/blog/feynman-open-source-ai-research-agent/ language: en description: "Evaluate Feynman, the open-source AI research agent: workflows, local setup, citation checks, GPU replication, costs, risks and team pilot plan." image: "https://wavect.io/img/blog/headers/header_feynman-open-source-ai-research-agent.png" --- [**Back**](/blog/overview/) [![Kevin Riedl](/img/team/kevin.webp)](/team/kevin-riedl/) [Kevin Riedl](/team/kevin-riedl/) https://linkedin.com/in/wsdt 11 min read · 3 Sep 2026 Last reviewed September 3, 2026 [**Next**](/blog/why-wavect-joined-goodfirms/) # Feynman Review: Is This Open-Source AI Research Agent Ready for Teams? TL;DR Feynman is an MIT-licensed, terminal-first AI research system that combines paper and web discovery, four specialist agents, source verification, paper-to-code audits, replication planning, optional compute execution and recurring research watches. It is unusually complete for a young open-source research tool, but it does not make research free, private or correct by default. A fully local setup requires a local model and carefully chosen local or keyless retrieval paths; hosted model, search and GPU providers create usage cost and data-governance work. Citation verification checks whether a source supports a claim and whether its URL resolves, but human domain review still decides whether the evidence is sufficient. Pilot Feynman on one bounded, repeatable research question and measure source precision, citation support, reviewer corrections, time to accepted brief, model and compute cost, and reproducibility before wider adoption. **Feynman is one of the most complete open-source attempts to turn deep research into an auditable terminal workflow.** Give it a topic and it can search papers, websites and repositories, delegate work to four specialist agents, draft a sourced brief, check citations, compare a paper with its code, plan a replication and monitor a topic over time. Our verdict after reviewing the public code and documentation on 3 September 2026: **Feynman is a strong pilot candidate for technical research teams, but “fully local” describes a possible configuration, not every default workflow.** Model calls, search providers and cloud GPUs can still move data outside your machine and create usage cost. Citation checks reduce a dangerous failure mode, but they do not replace a scientist, engineer or analyst who can judge whether the evidence is sufficient. This page owns the product-specific “Feynman review” and “open-source AI research agent” decision. Our [LangChain Deep Agents review](/blog/langchain-deep-agents-review/) covers a general-purpose agent harness, [Agent Reach](/blog/agent-reach-open-source-review/) covers low-cost source access, and [LLM-as-a-Verifier](/blog/llm-as-a-verifier/) covers fine-grained trajectory scoring. Keeping those jobs separate prevents a branded tool review from competing with broader architecture guides. ## What is Feynman? **Feynman is an MIT-licensed research-first CLI built on the Pi agent runtime and alphaXiv.** Its [official README](https://github.com/companion-inc/feynman/blob/main/README.md) describes research workflows for deep investigation, literature review, paper ranking, drafting, peer-style review, code audit, replication, ML recipes and recurring watches. Outputs are written to local project folders, with provenance artifacts for source-heavy workflows. | Agent | Primary job | What still needs human judgment | | --- | --- | --- | | Researcher | Find papers, web pages, repositories, datasets and prior art | Whether the search space and exclusion criteria are appropriate | | Reviewer | Critique drafts with severity-graded feedback and a revision plan | Whether the review standard matches your field or business decision | | Writer | Turn notes and evidence into structured reports or paper-style drafts | Whether the framing, uncertainty and recommendations are defensible | | Verifier | Check source URLs, citation support and paper-to-code claims | Whether the source itself is trustworthy and the evidence is sufficient | The four-agent story is useful, but the workflow is more important than the headcount. Feynman stores plans, intermediate research, drafts, verification notes and final outputs as files. That gives a team reviewable handoff points and a recovery path when a long run fails. It also means sensitive research notes need the same access, retention and backup policy as any other working dataset. ## How does Feynman deep research work? The project does not immediately unleash a swarm. Its [deep research workflow](https://feynman.is/docs/workflows/deep-research) starts with a written plan and waits for approval. Narrow explainers remain lead-owned; broader surveys can delegate parallel searches. The system then extracts findings, synthesizes agreement and disagreement, and runs a final claim-to-source verification pass. 1. **Plan the question.** Define subquestions, source strategy, task ledger and verification log before spending model budget. 2. **Choose the scale.** Keep a narrow question with one lead agent, or parallelize a broad evidence search. 3. **Extract, do not merely summarize.** Capture methods, results, limitations and claim locations from the strongest sources. 4. **Synthesize conflicts.** Separate consensus, disagreement and unanswered questions instead of forcing one neat answer. 5. **Verify before delivery.** Reopen cited sources, test whether they support the attached claim, and record gaps. This plan gate is good product design. It makes scope and source standards visible before the expensive part begins. The quality still depends on the prompt, the selected model, retrieval coverage and the reviewer who approves the plan. ## What can Feynman do beyond a literature review? | Workflow | Business or research decision it supports | Important boundary | | --- | --- | --- | | `deepresearch` | Build a sourced brief across papers, web sources and code | Retrieval coverage is not guaranteed | | `lit` and `rank` | Map a field and decide which papers deserve attention first | Ranking criteria still encode judgment | | `audit` | Check whether public code matches the paper's stated method | Only public and discoverable implementation evidence can be checked | | `replicate` | Turn a claim into requirements, steps, assumptions and success criteria | Execution requires an explicit environment choice | | `watch` | Track new papers, releases or evidence for a topic | Recurrence exists only when a scheduler tool is available | | `recipe` and `autoresearch` | Translate research into an implementable ML experiment loop | Compute, data rights and evaluation remain your responsibility | ## Is Feynman really fully local? **It can be configured for local model inference, but a normal research workflow still reaches external sources.** The [standalone installer documentation](https://feynman.is/docs/getting-started/installation) says the prebuilt bundle includes a pinned Node.js runtime and verifies its SHA-256 before replacement. An npm installation is also available and requires a compatible Node version. For supply-chain-sensitive teams, inspect the script, pin an exact release and verify the published checksum rather than treating a pipe-to-shell command as a policy exception. Feynman's [setup guide](https://feynman.is/docs/getting-started/setup) supports hosted providers, Amazon Bedrock, LM Studio, LiteLLM, Ollama and vLLM. A local model can keep prompts and inference on infrastructure you control. Paper search, web retrieval, alphaXiv, optional hosted parsers and cloud compute still cross a network boundary unless you deliberately replace or disable them. Use a simple data-flow test before calling the system local: - Where does the research prompt go? - Which provider receives URLs, documents, repository content or search queries? - Which fetched pages are cached, where, and for how long? - Which credentials and OAuth tokens are stored under the Feynman home directory? - Does telemetry leave the machine, and does your policy allow its metadata? “Runs from my terminal” is an interface statement. “All data stays on my machine” is an architecture claim. Verify the second one from the exact provider and search configuration you deploy. ## Does Feynman eliminate hallucinated citations? **No, but it creates a much better failure boundary than citation-shaped text with no verification step.** The verifier is instructed to fetch each URL, locate the supporting passage, classify the claim as supported, overstated, contradicted or unsupported, remove claims it cannot source, and avoid fake certainty. That is the right direction. Three limits remain. A live URL can contain weak evidence. A source can support the words while the model misses a methodological flaw. Search can omit the paper that changes the conclusion. Treat the final brief as an auditable draft, not an autonomous sign-off. Our [AI agent evaluation and sandbox checklist](/blog/ai-agent-eval-sandbox-security-checklist/) shows how to add deterministic checks, adversarial cases and human approval around agent output. ## Can Feynman audit and replicate papers? The [code audit workflow](https://feynman.is/docs/workflows/audit) extracts concrete paper claims, then searches the associated public repository for matching configuration, training and evaluation behavior. It reports mismatches, missing implementations and reproducibility risks with file locations. That is valuable due diligence before a team builds on a published result, but it cannot inspect private training code, unavailable data or undocumented operator choices. The [replication workflow](https://feynman.is/docs/workflows/replication) creates requirements, an experimental recipe, underspecified details, risks and success criteria. It executes only after the user chooses local, isolated, Docker, Modal, RunPod or plan-only operation. This approval boundary matters because a replication can download untrusted code, process licensed data and spend real GPU budget. Use the word “replicated” only when the planned checks passed against preserved scripts and raw outputs. A plausible plan is not a reproduction, and a matching headline metric is not enough if the dataset split or evaluation path changed. ## Do recurring research watches run automatically? Only when scheduling capability is actually available. The [watch documentation](https://feynman.is/docs/workflows/watch) says Feynman always creates a baseline and follow-up plan, but schedules recurrence only when its scheduling tool is visible. Otherwise it records the gap and gives you a refresh prompt. That is honest behavior and an important operational detail for teams expecting continuous monitoring. ## What does Feynman cost? The software is [MIT licensed](https://github.com/companion-inc/feynman/blob/main/LICENSE), so there is no licence fee. The total cost of an accepted research output has at least five other lines: - **Model inference:** lead-agent reasoning, parallel researchers, review, writing and verification all consume tokens or local compute. - **Search and extraction:** some providers are keyless, while hosted search, parsing and grounding services may charge by request or result. - **Experiment compute:** local GPU time, cloud GPU minutes, storage and data transfer can dominate a replication. - **Human review:** a domain expert must resolve uncertain evidence, unsupported implications and risky recommendations. - **Operations:** version pinning, credentials, retention, monitoring, provider changes and failed runs require ownership. Measure cost per accepted brief or verified experiment, not cost per model call. A cheap report that takes two senior hours to repair is expensive. A more costly run that exposes disagreement early may be the better system. ## Who should adopt Feynman? | Team | Fit | Why | | --- | --- | --- | | ML or R&D team reviewing papers and code | Strong pilot | Audit, ranking, replication and provenance match the work | | Technical due-diligence team | Strong pilot | Source-backed briefs and code checks create reviewable evidence | | Product team tracking a fast-moving technical field | Good fit with operations | Watches and reusable workflows reduce repeated setup | | Student needing a one-off summary | Possible overkill | The terminal, provider setup and review workflow add friction | | Regulated team expecting autonomous approval | Poor fit without controls | Human accountability, data governance and validation cannot be delegated | | Team needing a general production agent harness | Choose another category | Feynman is deliberately research-first, not a generic business automation platform | ## A two-week Feynman pilot that produces a buying decision 1. **Choose one repeatable research question.** Use a topic where your team already has a trusted human brief and known must-find sources. 2. **Freeze the acceptance rubric.** Score recall of must-find evidence, irrelevant-source rate, claim support, reviewer corrections and time to approval. 3. **Pin the release and model.** Record the Feynman version, model, search route, prompts and retrieval date so runs can be compared. 4. **Start without experiment execution.** Validate discovery, synthesis and citation support before granting code or GPU access. 5. **Run one adversarial topic.** Include conflicting papers, a dead link, a paper with incomplete code and an attractive unsupported claim. 6. **Review the data flow.** Map every hosted provider, credential, cache, retention window and outbound request. 7. **Test one paper audit.** Have an engineer verify a sample of reported matches and mismatches directly in the repository. 8. **Test one bounded replication plan.** Compare its requirements and assumptions with a domain expert before any compute spend. 9. **Count total cost.** Include tokens, search, GPUs, failed runs and reviewer minutes. 10. **Set the adoption gate in advance.** Expand only if accepted-brief time or evidence coverage improves without lowering citation support or control. ## Frequently asked questions ### What is Feynman AI? Feynman is an open-source, terminal-first AI research system. It searches papers, web sources and repositories, coordinates researcher, reviewer, writer and verifier agents, and writes sourced research artifacts to local project folders. ### Is Feynman open source and free? The code is MIT licensed, so there is no licence fee. Model APIs, hosted search, document parsing, cloud GPUs, storage, operations and human review can still create substantial cost. ### Can Feynman run fully locally? It supports local model servers such as LM Studio, Ollama and vLLM. A truly local data path also requires you to avoid or self-host external search, parsing and compute services. Running from a terminal alone does not guarantee that data stays local. ### Does Feynman verify every citation? Its verifier is designed to fetch source URLs and test whether they support the attached claims, removing unsupported statements when necessary. This reduces citation errors but does not establish the quality of the source, complete search recall or domain-level validity. ### Can Feynman run paper experiments on GPUs? It can plan replications and, after an explicit environment choice, help execute them locally, in isolation, with Docker, or through documented cloud GPU paths. Data rights, code safety, budget and scientific acceptance remain the user's responsibility. ### Is Feynman ready for enterprise research? It is ready for a bounded, measured pilot. Wider production use needs version pinning, provider review, credential controls, data classification, deterministic evaluation, cost limits, audit logs and accountable human approval. ## Research boundary *Status checked 3 September 2026 against Feynman's public repository, license and official workflow documentation. This is an independent architecture and buyer review, not a sponsored post, security audit or controlled benchmark. We did not install Feynman, submit private documents, connect paid providers or execute a GPU replication. The project is moving quickly, so pin a release and recheck its current documentation before adoption.* ## Final thoughts Feynman's strongest idea is not that four agents can write a report. It is that research should leave a visible chain from plan to sources, synthesis, verification and, when needed, experimental artifacts. That chain gives a technical team something concrete to review instead of asking whether polished prose feels convincing. The honest buying decision is conditional. Pilot Feynman when repeated technical research is expensive and source accountability matters. Keep humans responsible for scope, evidence quality and approval. Call the system fully local only after you map its complete data path. Then judge it by accepted research outcomes, citation support, reproducibility and total review cost, not by the speed of the first impressive brief. ## You may also like.. [**Agent Reach Review: Costs, Security and Real Limits** Separate the source-access layer from the research workflow that interprets it.](/blog/agent-reach-open-source-review/) [**LangChain Deep Agents Review** Compare a research-specific agent system with a general-purpose production harness.](/blog/langchain-deep-agents-review/) Agent engineering ## Continue through this cluster Coding agents, MCP, context systems, evaluation and the controls required for dependable automation. [Start with the cornerstone**Graph Engineering for AI Agents: When Does a Knowledge Graph Pay Off?**](/blog/graph-engineering-ai-agents/) - [Obscura Browser Review: Claims, Limits and Production Fit](/blog/obscura-rust-browser-ai-agents/) - [Claude Code Design System: 4 Parts for On-Brand UI](/blog/claude-code-design-system-files/) - [ChatGPT Can Now Log In Without Seeing Your Password](/blog/chatgpt-cloud-browser-secure-login/) - [AI Agent Harness, Explained: The Reliability Layer Around an LLM](/blog/agent-harness-engineering/) - [Why Agent Edits Need Semantic Identity: Building SEMAPRAX in Rust](/blog/semantic-identity-rust-agent-edits/) [**Back**](/blog/overview/) [![Kevin Riedl](/img/team/kevin.webp)](/team/kevin-riedl/) [Kevin Riedl](/team/kevin-riedl/) https://linkedin.com/in/wsdt 11 min read · 3 Sep 2026 Last reviewed September 3, 2026 [**Next**](/blog/why-wavect-joined-goodfirms/) ## Structured Data ```json { "@context": "https://schema.org", "@graph": [ { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ], "employee": [ { "@id": "https://wavect.io/team/kevin-riedl/#person", "@type": "Person", "jobTitle": "Managing Director", "name": "Kevin Riedl", "url": "https://wavect.io/team/kevin-riedl/", "worksFor": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] } }, { "@id": "https://wavect.io/team/christof-jori/#person", "@type": "Person", "jobTitle": "Managing Director", "name": "Christof Jori", "url": "https://wavect.io/team/christof-jori/", "worksFor": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] } } ], "founder": [ { "@id": "https://wavect.io/team/kevin-riedl/#person", "@type": "Person", "jobTitle": "Managing Director", "name": "Kevin Riedl", "url": "https://wavect.io/team/kevin-riedl/", "worksFor": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] } }, { "@id": "https://wavect.io/team/christof-jori/#person", "@type": "Person", "jobTitle": "Managing Director", "name": "Christof Jori", "url": "https://wavect.io/team/christof-jori/", "worksFor": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] } } ], "legalRepresentative": [ { "@id": "https://wavect.io/team/kevin-riedl/#person", "@type": "Person", "jobTitle": "Managing Director", "name": "Kevin Riedl", "url": "https://wavect.io/team/kevin-riedl/", "worksFor": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] } }, { "@id": "https://wavect.io/team/christof-jori/#person", "@type": "Person", "jobTitle": "Managing Director", "name": "Christof Jori", "url": "https://wavect.io/team/christof-jori/", "worksFor": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] } } ], "name": "Wavect GmbH", "subjectOf": { "@id": "https://wavect.io/verified-claims.json#dataset", "@type": "Dataset", "creator": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] }, "description": "A machine-readable registry of quantitative and qualitative claims published by Wavect, with review dates, localized page appearances and public third-party citations where available.", "inLanguage": "en", "isAccessibleForFree": true, "license": "https://creativecommons.org/licenses/by/4.0/", "name": "Wavect verified publication claims", "url": "https://wavect.io/verified-claims.json" }, "url": "https://wavect.io/" }, { "@id": "https://wavect.io/team/kevin-riedl/#person", "@type": "Person", "jobTitle": "Managing Director", "name": "Kevin Riedl", "sameAs": [ "https://www.wikidata.org/wiki/Q139796365", "https://www.linkedin.com/in/wsdt", "https://github.com/wsdt" ], "url": "https://wavect.io/team/kevin-riedl/", "worksFor": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] } }, { "@id": "https://wavect.io/team/christof-jori/#person", "@type": "Person", "jobTitle": "Managing Director", "name": "Christof Jori", "sameAs": [ "https://www.wikidata.org/wiki/Q139796367", "https://www.linkedin.com/in/jocr77/", "https://github.com/jo-chris" ], "url": "https://wavect.io/team/christof-jori/", "worksFor": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] } }, { "@id": "https://wavect.io/#website", "@type": "WebSite", "inLanguage": [ "en", "de", "es", "zh" ], "name": "Wavect", "potentialAction": { "@type": "SearchAction", "query-input": "required name=search_term_string", "target": { "@type": "EntryPoint", "urlTemplate": "https://wavect.io/search/?q={search_term_string}" } }, "publisher": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] }, "url": "https://wavect.io/" }, { "@id": "https://wavect.io/blog/feynman-open-source-ai-research-agent/#webpage", "@type": "WebPage", "dateModified": "2026-09-03", "inLanguage": "en", "isPartOf": { "@id": "https://wavect.io/#website", "@type": "WebSite" }, "lastReviewed": "2026-09-03", "url": "https://wavect.io/blog/feynman-open-source-ai-research-agent/" } ] } ``` ```json { "@context": "https://schema.org", "@type": "BlogPosting", "abstract": "Feynman is an MIT-licensed, terminal-first AI research system that combines paper and web discovery, four specialist agents, source verification, paper-to-code audits, replication planning, optional compute execution and recurring research watches. It is unusually complete for a young open-source research tool, but it does not make research free, private or correct by default. A fully local setup requires a local model and carefully chosen local or keyless retrieval paths; hosted model, search and GPU providers create usage cost and data-governance work. Citation verification checks whether a source supports a claim and whether its URL resolves, but human domain review still decides whether the evidence is sufficient. Pilot Feynman on one bounded, repeatable research question and measure source precision, citation support, reviewer corrections, time to accepted brief, model and compute cost, and reproducibility before wider adoption.", "articleBody": " Blog overview/AI and agents/Agent engineering Feynman Review: Is This Open-Source AI Research Agent Ready for Teams? TL;DR Feynman is an MIT-licensed, terminal-first AI research system that combines paper and web discovery, four specialist agents, source verification, paper-to-code audits, replication planning, optional compute execution and recurring research watches. It is unusually complete for a young open-source research tool, but it does not make research free, private or correct by default. A fully local setup requires a local model and carefully chosen local or keyless retrieval paths; hosted model, search and GPU providers create usage cost and data-governance work. Citation verification checks whether a source supports a claim and whether its URL resolves, but human domain review still decides whether the evidence is sufficient. Pilot Feynman on one bounded, repeatable research question and measure source precision, citation support, reviewer corrections, time to accepted brief, model and compute cost, and reproducibility before wider adoption. Feynman is one of the most complete open-source attempts to turn deep research into an auditable terminal workflow. Give it a topic and it can search papers, websites and repositories, delegate work to four specialist agents, draft a sourced brief, check citations, compare a paper with its code, plan a replication and monitor a topic over time. Our verdict after reviewing the public code and documentation on 3 September 2026: Feynman is a strong pilot candidate for technical research teams, but “fully local” describes a possible configuration, not every default workflow. Model calls, search providers and cloud GPUs can still move data outside your machine and create usage cost. Citation checks reduce a dangerous failure mode, but they do not replace a scientist, engineer or analyst who can judge whether the evidence is sufficient. This page owns the product-specific “Feynman review” and “open-source AI research agent” decision. Our LangChain Deep Agents review covers a general-purpose agent harness, Agent Reach covers low-cost source access, and LLM-as-a-Verifier covers fine-grained trajectory scoring. Keeping those jobs separate prevents a branded tool review from competing with broader architecture guides. What is Feynman? Feynman is an MIT-licensed research-first CLI built on the Pi agent runtime and alphaXiv. Its official README describes research workflows for deep investigation, literature review, paper ranking, drafting, peer-style review, code audit, replication, ML recipes and recurring watches. Outputs are written to local project folders, with provenance artifacts for source-heavy workflows. AgentPrimary jobWhat still needs human judgment ResearcherFind papers, web pages, repositories, datasets and prior artWhether the search space and exclusion criteria are appropriate ReviewerCritique drafts with severity-graded feedback and a revision planWhether the review standard matches your field or business decision WriterTurn notes and evidence into structured reports or paper-style draftsWhether the framing, uncertainty and recommendations are defensible VerifierCheck source URLs, citation support and paper-to-code claimsWhether the source itself is trustworthy and the evidence is sufficient The four-agent story is useful, but the workflow is more important than the headcount. Feynman stores plans, intermediate research, drafts, verification notes and final outputs as files. That gives a team reviewable handoff points and a recovery path when a long run fails. It also means sensitive research notes need the same access, retention and backup policy as any other working dataset. How does Feynman deep research work? The project does not immediately unleash a swarm. Its deep research workflow starts with a written plan and waits for approval. Narrow explainers remain lead-owned; broader surveys can delegate parallel searches. The system then extracts findings, synthesizes agreement and disagreement, and runs a final claim-to-source verification pass. Plan the question. Define subquestions, source strategy, task ledger and verification log before spending model budget. Choose the scale. Keep a narrow question with one lead agent, or parallelize a broad evidence search. Extract, do not merely summarize. Capture methods, results, limitations and claim locations from the strongest sources. Synthesize conflicts. Separate consensus, disagreement and unanswered questions instead of forcing one neat answer. Verify before delivery. Reopen cited sources, test whether they support the attached claim, and record gaps. This plan gate is good product design. It makes scope and source standards visible before the expensive part begins. The quality still depends on the prompt, the selected model, retrieval coverage and the reviewer who approves the plan. What can Feynman do beyond a literature review? WorkflowBusiness or research decision it supportsImportant boundary", "articleSection": "AI Agents", "author": { "@id": "https://wavect.io/team/kevin-riedl/#person", "@type": "Person", "name": "Kevin Riedl", "sameAs": [ "https://www.wikidata.org/wiki/Q139796365", "https://www.linkedin.com/in/wsdt", "https://github.com/wsdt" ], "url": "https://wavect.io/team/kevin-riedl/" }, "citation": [ { "@type": "WebPage", "name": "official README", "url": "https://github.com/companion-inc/feynman/blob/main/README.md" }, { "@type": "WebPage", "name": "deep research workflow", "url": "https://feynman.is/docs/workflows/deep-research" }, { "@type": "WebPage", "name": "standalone installer documentation", "url": "https://feynman.is/docs/getting-started/installation" }, { "@type": "WebPage", "name": "setup guide", "url": "https://feynman.is/docs/getting-started/setup" }, { "@type": "WebPage", "name": "code audit workflow", "url": "https://feynman.is/docs/workflows/audit" }, { "@type": "WebPage", "name": "replication workflow", "url": "https://feynman.is/docs/workflows/replication" }, { "@type": "WebPage", "name": "watch documentation", "url": "https://feynman.is/docs/workflows/watch" }, { "@type": "WebPage", "name": "MIT licensed", "url": "https://github.com/companion-inc/feynman/blob/main/LICENSE" } ], "dateModified": "2026-09-03", "datePublished": "2026-09-03", "description": "Feynman is an MIT-licensed, terminal-first AI research system that combines paper and web discovery, four specialist agents, source verification, paper-to-code audits, replication planning, optional compute execution and recurring research watches. It is unusually complete for a young open-source research tool, but it does not make research free, private or correct by default. A fully local setup requires a local model and carefully chosen local or keyless retrieval paths; hosted model, search and GPU providers create usage cost and data-governance work. Citation verification checks whether a source supports a claim and whether its URL resolves, but human domain review still decides whether the evidence is sufficient. Pilot Feynman on one bounded, repeatable research question and measure source precision, citation support, reviewer corrections, time to accepted brief, model and compute cost, and reproducibility before wider adoption.", "headline": "Feynman Review: Is This Open-Source AI Research Agent Ready for Teams?", "image": "https://wavect.io/img/blog/headers/header_feynman-open-source-ai-research-agent.svg", "inLanguage": "en", "keywords": "AI Agents, Open Source", "mainEntityOfPage": { "@id": "https://wavect.io/blog/feynman-open-source-ai-research-agent/", "@type": "WebPage" }, "publisher": { "@id": "https://wavect.io/#organization", "@type": [ "Organization", "ProfessionalService", "LocalBusiness" ] }, "url": "https://wavect.io/blog/feynman-open-source-ai-research-agent/", "wordCount": 2503 } ``` ```json { "@context": "https://schema.org", "@type": "BreadcrumbList", "itemListElement": [ { "@type": "ListItem", "item": "https://wavect.io/", "name": "Home", "position": 1 }, { "@type": "ListItem", "item": "https://wavect.io/blog/overview/", "name": "Blog overview", "position": 2 }, { "@type": "ListItem", "item": "https://wavect.io/blog/topics/ai-agents/", "name": "AI and agents", "position": 3 }, { "@type": "ListItem", "item": "https://wavect.io/blog/clusters/agent-engineering/", "name": "Agent engineering", "position": 4 }, { "@type": "ListItem", "item": "https://wavect.io/blog/feynman-open-source-ai-research-agent/", "name": "Feynman AI Research Agent Review: Costs & Fit", "position": 5 } ] } ``` ```json { "@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [ { "@type": "Question", "acceptedAnswer": { "@type": "Answer", "text": "Feynman is an open-source, terminal-first AI research system. It searches papers, web sources and repositories, coordinates researcher, reviewer, writer and verifier agents, and writes sourced research artifacts to local project folders." }, "name": "What is Feynman AI?" }, { "@type": "Question", "acceptedAnswer": { "@type": "Answer", "text": "The code is MIT licensed, so there is no licence fee. Model APIs, hosted search, document parsing, cloud GPUs, storage, operations and human review can still create substantial cost." }, "name": "Is Feynman open source and free?" }, { "@type": "Question", "acceptedAnswer": { "@type": "Answer", "text": "It supports local model servers such as LM Studio, Ollama and vLLM. A truly local data path also requires you to avoid or self-host external search, parsing and compute services. Running from a terminal alone does not guarantee that data stays local." }, "name": "Can Feynman run fully locally?" }, { "@type": "Question", "acceptedAnswer": { "@type": "Answer", "text": "Its verifier is designed to fetch source URLs and test whether they support the attached claims, removing unsupported statements when necessary. This reduces citation errors but does not establish the quality of the source, complete search recall or domain-level validity." }, "name": "Does Feynman verify every citation?" }, { "@type": "Question", "acceptedAnswer": { "@type": "Answer", "text": "It can plan replications and, after an explicit environment choice, help execute them locally, in isolation, with Docker, or through documented cloud GPU paths. Data rights, code safety, budget and scientific acceptance remain the user's responsibility." }, "name": "Can Feynman run paper experiments on GPUs?" }, { "@type": "Question", "acceptedAnswer": { "@type": "Answer", "text": "It is ready for a bounded, measured pilot. Wider production use needs version pinning, provider review, credential controls, data classification, deterministic evaluation, cost limits, audit logs and accountable human approval." }, "name": "Is Feynman ready for enterprise research?" } ] } ```