PromptHub review
A prompt management platform for teams building LLM features, with Git-style versioning, side-by-side model testing, evaluations, CI pipelines and an API, free for public prompts.
By Noahlast updated pricing verified
PromptHub screenshots & video
PromptHub in 20 seconds
Best for: Versioning and testing the production prompts behind an LLM feature, with evaluation pipelines that stop a regressed prompt from shipping.
Pricing: Free for public prompts with 2,000 requests a month; Pro $12/month ($9 billed yearly) for one user, Team $20 per user per month ($15 billed yearly), Enterprise custom, plus model token costs.
- PromptHub's free plan includes unlimited team members and 2,000 LLM requests a month, but every prompt is public.
- PromptHub Pro costs $12/month, or $9/month billed yearly, for one user with private prompts and 10,000 requests a month.
- PromptHub Team costs $20 per user per month, or $15 billed yearly, for 50,000 requests and 500 prompt enhancements a month plus evaluations and pipelines.
- PromptHub's API returns the latest prompt commit from any branch (/head) or runs it with logging (/run).
- PromptHub lists SOC 2 compliance as in progress on its Enterprise plan.
Limitations: Free-plan prompts are public.; Model token costs are extra, and requests are capped monthly.; Evaluations and pipelines need the Team plan.
Top alternatives: PromptLayer, Langfuse, KingPrompt, AIPRM, PromptBase — full list
What is PromptHub?
PromptHub is a prompt management tool: a shared workspace where product and engineering teams write, version, test and deploy the prompts behind LLM applications, plus a community library of public prompts. Each prompt project keeps its model, parameters, variables and system message under version control, with branches for environments such as staging and production and merge requests that log every change for review. Testing runs in a playground that compares models side by side, in chat testing that can restart a conversation from any message, in batch runs over datasets, and through evaluators that score outputs with string rules or an LLM judge. On the Team plan, pipelines trigger those evaluators on commits, merge requests or API calls and can stop a version from shipping. Applications call /head to fetch the latest commit from a branch, or /run to execute it through PromptHub with logging. Requests use the customer's own model API keys and count against a monthly request quota.
PromptHub is a prompt management and testing platform from Tethered Software Inc. for teams that ship LLM features. Prompts live in projects with Git-style branches, commits and merge requests; a playground compares outputs from OpenAI, Anthropic, Azure, Google, Meta, Bedrock and Mistral models side by side; batch tests and LLM-as-a-judge evaluations run prompts over datasets; and pipelines can block a prompt version that fails its checks. Apps fetch or run the latest commit from any branch through a REST API, and prompts can also be deployed as Zapier actions or shareable forms. The free plan allows unlimited members and 2,000 requests a month, but every prompt is public; Pro costs $12/month ($9 billed yearly) for one user with private prompts, and Team $20 per user per month ($15 billed yearly). Users pay model token costs to the providers.
Standout feature: Git-style branches, commits and merge requests for prompts, with pipelines that run evaluators on every commit or merge request and block a version that misses its pass threshold.
Biggest weakness: Private prompts and team testing features are priced per user: the free plan makes every prompt public, Pro covers one user at $12/month, and evaluations and CI pipelines need Team at $20 per user per month, all on top of model token costs paid directly to OpenAI, Anthropic or other providers; SOC 2 is still listed as in progress.
PromptHub: pros and cons
Pros
- The free plan includes unlimited team members, version control, batch testing, prompt chains and forms.
- Branches, commits and merge requests log every change to a prompt's text, parameters and model for review.
- Team-plan pipelines run evaluators on commits or merge requests and block versions that fail, guarding against secret leaks, profanity and regressions.
- One playground compares outputs from OpenAI, Anthropic, Azure, Google, Meta, Bedrock and Mistral models side by side.
- Heidi Health, an AI medical scribe company, reports 3x faster prompt testing after moving off spreadsheets and Jupyter notebooks.
Cons
- All free-plan prompts are public; private prompts start at $12/month for a single user.
- Evaluations and pipelines require the Team plan at $20 per user per month ($15 billed yearly).
- Model token costs are billed separately by the providers, and every plan caps LLM requests per month (2,000, 10,000 or 50,000).
- Signs of slower upkeep: the footer's API documentation link returned a 404 in October 2026, and the product blog's latest post is dated October 23, 2025.
PromptHub pricing
Free tier · paid from $12/mo · verified from the vendor's pricing page
| Plan | Price | Notes |
|---|---|---|
| Free | Free | Unlimited team members and public prompts, no private prompts, 2,000 LLM requests and 10 prompt enhancements a month, limited API access. |
| Pro | $12/mo | Billed monthly, or $9/month billed yearly; single user, unlimited private prompts, 10,000 requests and 200 prompt enhancements a month, full API access. |
| Team | $20/mo | Per user, billed monthly, or $15/user/month billed yearly; unlimited members, 50,000 requests and 500 prompt enhancements a month, evaluations, pipelines, team permissions and private Slack support. |
| Enterprise | Custom | Custom; SSO and SAML, custom limits and roles, custom model support and dedicated support; SOC 2 listed as in progress. |
What reviewers say
No third-party review scores on file yet.