promptfoo

Promptfoo is an open-source LLM eval and AI security testing...

8.8/10 Strong

Active

Community free / Enterprise custom / On-Premise custom

Best plan

Use Community for local evals and small-team red teaming...

Risk: Red-team and vulnerability tests are only useful when...

Try promptfoo free

Editorial · no paid placements

Should you use it?

promptfoo is an open-source LLM eval and AI security testing platform. Pick it when red teaming, jailbreak tests, vulnerability scanning, guardrails, model security, MCP proxy risk, code scanning, and compliance evidence matter before launch. Compare Braintrust for eval operations, Portkey or Helicone for gateway control, and LangSmith for LangChain-native tracing.

Buy if Teams red-teaming LLM apps before launch
Pick Use Community for local evals and small-team red teaming, Enterprise when teams need managed cloud, collaboration, continuous monitoring, dashboards, SSO, API access, and support, and On-Premise when infrastructure control and data isolation are mandatory
Skip if Buyers who only need provider routing and cost logs

Plan guidance

What to buy

Best plan Use Community for local evals and small-team red teaming, Enterprise when teams need managed cloud, collaboration, continuous monitoring, dashboards, SSO, API access, and support, and On-Premise when infrastructure control and data isolation are mandatory

Watch: Red-team and vulnerability tests are only useful when...

Price range Community free / Enterprise custom / On-Premise custom

Free Forever

Upgrade only if Not for buyers who only need provider routing and cost logs

Red-team and vulnerability tests are only useful when...

Current pricing source: promptfoo pricing

Fit

Use it for this, skip it for that

Best for

Teams red-teaming LLM apps before launch
Developers who need local evals across models, prompts, and RAG pipelines
Security and compliance teams testing jailbreaks, data leakage, and unsafe behavior
Organizations that need managed cloud or on-prem deployment for AI security workflows

Avoid if

Buyers who only need provider routing and cost logs
Teams with no owner for remediation and false-positive review
Simple prototypes that do not yet have a real target app
Buyers who need a general workflow automation canvas

Watch out: Red-team and vulnerability tests are only useful when mapped to real target apps, owned remediation workflows, false-positive review, compliance evidence, and repeatable release gates.

Recent changes

Only what affects the decision

Jun 28, 2026
Community
Listed as open source for individual developers and small teams, with LLM evaluation features, integrations, 10k red-team probes/month, local or self-hosted use, vulnerability scanning, and...
promptfoo pricing
Jun 28, 2026
Enterprise
Adds custom red-team limits, team sharing, continuous monitoring, centralized security/compliance dashboard, attack profiles, SSO, API access, managed cloud, professional services, priority...
promptfoo pricing
Jun 28, 2026
On-Premise
Listed for teams that need full infrastructure control, all Enterprise features, deployment on owned infrastructure, data isolation, dedicated runner, and assigned deployment engineer
promptfoo pricing

Alternatives

Best swaps

Hugging Face

Open AI collaboration hub for models, datasets, Spaces, inference endpoints, evaluations, and enterprise ML workflows.

Free hub access; Pro $9/mo; Team $20/user/mo; Enterprise from $50/user/mo; paid compute/storage · 9.3/10 LiteLLM

Open-source LLM gateway and Python SDK for one OpenAI-compatible interface across 100+ model providers, with routing, virtual ke

Free MIT core outside enterprise directory; Enterprise custom · 8.8/10 LlamaIndex

Open-source framework and managed LlamaCloud stack for building LLM agents over private data, RAG, document parsing, extraction,

Framework free MIT / LlamaParse Free 10K credits / Starter $50/month / Pro $500/month / Enterprise custom · 8.5/10

Build comparison

Proof and score math Verified Jun 28

Proof

Why this recommendation is trusted

Evidence promptfoo pricing

Source: Registered source
Freshness: Current
Confidence: High confidence
Verified: Jun 28, 2026
Review: Jul 12, 2026
Volatility: Volatile

High-volatility evidence needs frequent review.

Editorial score

Unweighted average of 4 axes · confidence high

Utility 9/10

How much real work it can do for a competent operator, end to end.
Value 9/10

What you get for the dollar relative to the closest alternative.
Moat 8/10

How hard it would be for a competitor to replicate the underlying advantage.
Longevity 9/10

How likely the product is to still be best-in-class 24 months out.

Verified facts

Best For Teams testing LLM applications for eval quality, red-team vulnerabilities, guardrail behavior, model security issues, MCP proxy risk, code scanning, and compliance workflows.
high Drifts 2026-06-28 promptfoo pricing
Pricing Anchor promptfoo pricing lists Community as Free Forever, Enterprise as Custom, and On-Premise as Custom.
high Volatile 2026-06-28 promptfoo pricing
Watch Out For Red-team and vulnerability tests are only useful when mapped to real target apps, owned remediation workflows, false-positive review, compliance evidence, and repeatable release gates.
high Drifts 2026-06-28 promptfoo docs
Open Source Or Local The promptfoo repository is MIT licensed, and the Community plan is positioned as a free open-source tool that can run locally or self-host on the buyer's own infrastructure.
high Drifts 2026-06-28 promptfoo license
Acquisition Status promptfoo says it is now part of OpenAI, while OpenAI announced on March 9, 2026 that it would acquire promptfoo subject to customary closing conditions.
high Volatile 2026-06-28 OpenAI promptfoo acquisition announcement

Full review notes Long-form details, FAQ, and source history

promptfoo is an LLM evaluation and AI security testing platform. It is used for prompt and model evals, red teaming, vulnerability scanning, guardrail testing, model security, MCP proxy checks, code scanning, and continuous monitoring.

The important June 2026 context: promptfoo says it is now part of OpenAI, and OpenAI announced on March 9, 2026 that it would acquire promptfoo subject to customary closing conditions. The open-source project continues to matter because the Community path is still the entry point for local evals and small teams.

System Verdict

Pick promptfoo when failure testing is the job. It is strongest for teams that need local evals, red-team probes, jailbreak proxy review, and compliance-oriented reporting.

Skip it when routing is the main problem. Portkey or Helicone fit better when the buyer needs a live traffic gateway before model providers.

Best plan guidance: use Community for local evals and small-team testing. Use Enterprise when teams need collaboration, continuous monitoring, dashboards, SSO, API access, managed cloud, services, and support. Use On-Premise when data isolation and owned infrastructure are mandatory.

Key Facts


Core job	, guardrails proxy, code scanning
Community	Free Forever, open source, local or self-hosted
Community red teaming	10k probes/month
Enterprise	Custom, with collaboration, monitoring, dashboards, SSO, API access, managed cloud, support
On-Premise	Custom, with owned-infrastructure deployment and data isolation
License	MIT license for the repository
Company status	promptfoo says it is now part of OpenAI; OpenAI announced an acquisition agreement on March 9, 2026

When To Pick promptfoo

You need red-team coverage before launch. promptfoo is built for finding jailbreaks, unsafe outputs, data leakage, and app-specific vulnerabilities.
You need repeatable evals. Local and self-hosted evals help teams compare prompts, models, RAG outputs, and application behavior.
You need security workflows. Enterprise adds collaboration, continuous monitoring, centralized dashboards, attack profiles, SSO, API access, and support.
You are testing MCP exposure. MCP proxy and related security testing matter as agent tools become production surfaces.
You need open-source leverage. Community is the right starting point when a developer wants local runs before a managed or on-prem decision.

When To Pick Something Else

Eval operations and release workflow: Braintrust when datasets, experiments, human review, scores, and production monitoring are the main buying job.
Gateway and governance: Portkey when routing, keys, provider logs, guardrails, and policy enforcement need to sit in live traffic.
Open-source observability/gateway: Helicone when the team wants request logging, cost tracking, caching, failover, and gateway controls.
LangChain operations: LangSmith when LangGraph/LangChain traces and deployments need a first-party home.
App building: Dify or Flowise when the missing tool is a visual AI app builder, not a test harness.

Pricing

promptfoo pricing was checked on June 28, 2026 against the official pricing page.

Plan	Price	Included shape	Buyer fit
Community	Free Forever	Open-source, local or self-hosted, all LLM eval features, model providers, integrations, 10k red-team probes/month, vulnerability scanning	Developers, small teams, local evals
Enterprise	Custom	Custom red-team limits, collaboration, continuous monitoring, compliance dashboard, attack profiles, SSO, API access, managed cloud, support, SLA	Security and product teams operating shared AI apps
On-Premise	Custom	All Enterprise features, deployment on owned infrastructure, complete data isolation, dedicated runner, deployment engineer	Regulated teams and strict data-control environments

The practical buying advice: Community is enough to prove whether the test harness fits. Enterprise or On-Premise becomes relevant when test results need to become organizational evidence, not only developer output.

Failure Modes

Red-team probes can become theater. Tests need to match the real app, tools, data, policies, and user pathways.
False positives need triage. Security tests require owner review, remediation workflow, and retest standards.
Custom targets need maintenance. As the app changes, eval config and attack profiles need to change too.
OpenAI ownership changes procurement. Teams should verify support, roadmap, data handling, and contract path after the acquisition update.
Security testing is not runtime enforcement. Pair promptfoo with gateway, guardrail, permission, and monitoring controls where live traffic needs protection.

Methodology

This page was produced by the aipedia.wiki editorial pipeline. Scoring follows the four-dimension rubric at /about/scoring/ (Utility x Value x Moat x Longevity, unweighted average). Last verified 2026-06-28 against promptfoo pricing, docs, OpenAI acquisition material, the promptfoo acquisition update, and the promptfoo license.

FAQ

Is promptfoo free? Yes for Community. promptfoo lists Community as Free Forever and open source, with local or self-hosted use.

Is promptfoo part of OpenAI? promptfoo says it is now part of OpenAI. OpenAI announced on March 9, 2026 that it would acquire promptfoo subject to customary closing conditions.

promptfoo vs Braintrust? promptfoo is stronger for red teaming, vulnerability scanning, guardrails, and AI security testing. Braintrust is stronger for eval operations, datasets, experiments, scored outputs, and release-quality workflows.

Sources

promptfoo pricing: Community, Enterprise, On-Premise, red-team limits, collaboration, monitoring, and deployment options
promptfoo docs: product documentation for eval and testing workflows
promptfoo joining OpenAI update: promptfoo company update and open-source continuity statement
OpenAI promptfoo acquisition announcement: OpenAI acquisition announcement and closing-condition language
promptfoo license: MIT license

Category: AI Infrastructure · AI Coding · AI Automation
Alternatives: Braintrust · Portkey · Helicone · LangSmith

Share LinkedIn

Was this review helpful?

Embed this score on your site Free. Links back.

HTML

<a href="https://aipedia.wiki/tools/promptfoo/" target="_blank" rel="noopener"><img src="https://aipedia.wiki/badges/promptfoo.svg" alt="promptfoo on aipedia.wiki" width="260" height="72" /></a>

Markdown

[![promptfoo on aipedia.wiki](https://aipedia.wiki/badges/promptfoo.svg)](https://aipedia.wiki/tools/promptfoo/)

Badge value auto-updates if the editorial score changes. Attribution via the link is required.

Cite this page For journalists, researchers, and bloggers

News writers

According to aipedia.wiki Editorial at aipedia.wiki (https://aipedia.wiki/tools/promptfoo/)

APA

aipedia.wiki Editorial. (2026). promptfoo: Editorial Review. aipedia.wiki. Retrieved July 2, 2026, from https://aipedia.wiki/tools/promptfoo/

MLA 9

aipedia.wiki Editorial. "promptfoo: Editorial Review." aipedia.wiki, 2026, https://aipedia.wiki/tools/promptfoo/. Accessed July 2, 2026.

Chicago

aipedia.wiki Editorial. 2026. "promptfoo: Editorial Review." aipedia.wiki. https://aipedia.wiki/tools/promptfoo/.

BibTeX

@misc{promptfoo-editorial-review-2026,
  author = {{aipedia.wiki Editorial}},
  title = {promptfoo: Editorial Review},
  year = {2026},
  publisher = {aipedia.wiki},
  url = {https://aipedia.wiki/tools/promptfoo/},
  note = {Accessed: 2026-07-02}
}

Spotted an error or want to share your experience with promptfoo?

Every tool page is re-verified on a recurring cycle, and corrections land faster when readers flag them directly. If you spot a stale fact, a missing capability, or have used promptfoo and want to share what worked or didn't, the editorial desk reviews every message sent through this form.

Email editorial@aipedia.wiki

Report outdated info Help us keep this page accurate