Does Mushi replace PostHog session replay?

No. PostHog records every session; Mushi records the moment a user reports a bug, with the screenshot and console/network tail, and then writes the diagnosis. Keep PostHog for replay and product analytics; add Mushi when you want the report to arrive already explained.

Does PostHog fix errors with AI?

Yes, in two ways. Every error-tracking issue has a "Fix with AI" button that writes a fix or explain prompt for PostHog AI or a coding assistant, and PostHog AI can search errors and find their root cause. Self-driving, in open beta, turns recurring high-impact errors into Inbox reports and opens a draft pull request when a code fix is possible: 3 a month free, then $15 per PR. Both are on PostHog's free plan (500 AI credits and 3 PRs a month). The difference with Mushi is where it starts: PostHog works from the exceptions it captured; Mushi starts from what a user reported, with their screenshot, writes a plain-English diagnosis (50 a month on Free Cloud, unlimited self-hosted with your own key), and pulls the fix context into Cursor or Claude Code over MCP.

Can I run PostHog and Mushi together?

Yes. They do not conflict in the browser. There is no PostHog plugin for Mushi today, so a report will not deep-link to the PostHog replay; you can add the PostHog session id to Mushi metadata yourself.

Does Mushi do product analytics?

A basic version: Mushi.track() with funnels, paths, people and retention in the same console, 90-day retention by default. PostHog is the fuller analytics product; Mushi's is there so a solo builder does not need a second tool to see a funnel.

What do the free tiers cover?

PostHog (2026-09-21): 5K session recordings, 100K exceptions and 1M analytics events a month, 500 PostHog AI credits (worth $5) and 3 Inbox pull requests a month, no card, one project, one-year retention. Mushi: 50 diagnoses a month, no card; self-hosting is free.

PostHog session replay vs Mushi Mushi

PostHog records every session and every exception, and gives you product analytics on top. Mushi records the moment a user says something is wrong, attaches the screen they were looking at and the console/network tail, and writes the diagnosis. They answer different questions, and this page is mostly about which question you have.

Both now use AI on bugs. PostHog puts a “Fix with AI” prompt on every error-tracking issue, and its Self-driving beta opens a draft pull request for recurring high-impact errors, with a small free allowance each month and a price per PR after that (both in the table below). That path starts from the exceptions PostHog captured. Mushi starts from what a user reported and the screen they were on, writes a plain-English diagnosis on every plan, and hands the fix context to Cursor or Claude Code over MCP.

Facts checked 2026-09-21 against the linked sources. Next review due 2026-12-21. Rows marked “unverified — check before publishing” could not be confirmed on that date.

What each one is

PostHog: A product-analytics platform with session replay, error tracking, feature flags, surveys and more, priced per product with a monthly free allowance on each.

Claim	PostHog	Source
Session replay free tier	5K recordings per month	checked 2026-09-21
Error tracking free tier	100K exceptions per month	checked 2026-09-21
Analytics free tier	1M events per month	checked 2026-09-21
Free plan limits	No credit card; 1 project; 1-year data retention; usage stops at the free limits	checked 2026-09-21
Open source	An MIT-licensed open-source product is available (pricing page FAQ)	checked 2026-09-21
AI free tier	500 PostHog AI credits per month (worth $5) and 3 pull requests per month from the Inbox (beta)	checked 2026-09-21
AI on errors	PostHog AI can search errors and find their root cause in the web app. Every error-tracking issue has a "Fix with AI" button that writes a fix or explain prompt for PostHog AI or a coding assistant.	checked 2026-09-21
AI pull requests	Self-driving (open beta) turns recurring high-impact errors into Inbox reports and, when a code fix is possible, opens a draft pull request for you to review. The first 3 PRs each month are free, then $15 per PR.	checked 2026-09-21

Mushi Mushi: An in-app bug reporter for production apps: a user reports from the widget, Mushi attaches the screenshot and console/network tail, writes a plain-English diagnosis, and hands a fix prompt to Cursor or Claude Code over MCP.

Claim	Mushi Mushi	Source
Free tier	Free Cloud: 50 diagnoses per month (hard stop; later reports are kept, not diagnosed), 1 seat, 7-day retention, no card	checked 2026-09-21
Paid tiers	Indie $15/mo (500 diagnoses) · Pro $49/mo (2,000 diagnoses)	checked 2026-09-21
License	SDKs MIT; server AGPLv3	checked 2026-09-21
Self-host	Free, whole stack, one command onto your own Supabase project (free tier works); bring your own Anthropic key (SELF_HOSTED.md)	checked 2026-09-21
Editor loop	MCP server for Cursor / Claude Code; no second LLM key in the MCP process	checked 2026-09-21
Diagnosis accuracy	An LLM reading of the report and your repo. It can be wrong or name the wrong file; read it before you apply the fix.	checked 2026-09-21

Side by side

Question	PostHog	Mushi
What it captures	Every session as a replay, every exception as an event, plus product analytics	The sessions where a user said something was wrong, with the screenshot and console/network tail
Where the bug is found	You read the exception and the replay, or ask PostHog AI to find the root cause	Mushi writes the diagnosis: what broke, why, which files
Bugs that never throw	Visible in replay if you know which session to open	The user tells you, and the report carries the screen they were looking at
Editor hand-off	"Fix with AI" on each error issue writes a fix prompt for PostHog AI or your coding assistant	MCP: diagnosis, fix context and lessons in Cursor / Claude Code
Draft pull requests	Self-driving (open beta) opens one for recurring high-impact errors: 3 free a month, then $15 per PR	Optional agent-drafted PR from any report
Product analytics	Full suite: 1M events/mo free	Basic: track(), funnels, paths, people, retention in the same console
Free tier	5K replays + 100K exceptions/mo, 500 AI credits, 3 AI PRs, no card, 1 project	50 diagnoses/mo, 1 seat, 7-day retention, no card
Self-host	MIT open-source product available	Free, AGPLv3 server, MIT SDKs

They work alongside each other

Nothing here conflicts in the browser: the PostHog snippet and the Mushi SDK run side by side. There is no PostHog plugin for Mushi today, so a Mushi report will not deep-link to the matching replay. If you want that, add the PostHog session id to Mushi’s metadata when you initialise the SDK and the report will carry it.

Mushi also ships a basic product-analytics surface of its own, Mushi.track() with funnels, paths, people and retention in the same console. It exists so a solo builder does not need a second tool to see a signup funnel; it is not a PostHog replacement and does not try to be. See the SDK analytics reference.

Pick PostHog if

You want replay for every session, not only reported ones, and product analytics, flags and surveys in the same place.
Your bugs are found by watching replays rather than by users telling you.
You want AI fix prompts, and beta draft pull requests, on the exceptions PostHog already captures.

Pick Mushi if

You want the bug to start from what the user said and saw, and to arrive explained, with the screenshot and a fix prompt.
You fix bugs in Cursor or Claude Code and want the report in the editor.
You want the diagnosis on every plan, including self-hosted with your own LLM key.

Start free — 50 diagnoses/mo, no card

Frequently asked questions

Does Mushi replace PostHog session replay?: No. PostHog records every session; Mushi records the moment a user reports a bug, with the screenshot and console/network tail, and then writes the diagnosis. Keep PostHog for replay and product analytics; add Mushi when you want the report to arrive already explained.
Does PostHog fix errors with AI?: Yes, in two ways. Every error-tracking issue has a "Fix with AI" button that writes a fix or explain prompt for PostHog AI or a coding assistant, and PostHog AI can search errors and find their root cause. Self-driving, in open beta, turns recurring high-impact errors into Inbox reports and opens a draft pull request when a code fix is possible: 3 a month free, then $15 per PR. Both are on PostHog's free plan (500 AI credits and 3 PRs a month). The difference with Mushi is where it starts: PostHog works from the exceptions it captured; Mushi starts from what a user reported, with their screenshot, writes a plain-English diagnosis (50 a month on Free Cloud, unlimited self-hosted with your own key), and pulls the fix context into Cursor or Claude Code over MCP.
Can I run PostHog and Mushi together?: Yes. They do not conflict in the browser. There is no PostHog plugin for Mushi today, so a report will not deep-link to the PostHog replay; you can add the PostHog session id to Mushi metadata yourself.
Does Mushi do product analytics?: A basic version: Mushi.track() with funnels, paths, people and retention in the same console, 90-day retention by default. PostHog is the fuller analytics product; Mushi's is there so a solo builder does not need a second tool to see a funnel.
What do the free tiers cover?: PostHog (2026-09-21): 5K session recordings, 100K exceptions and 1M analytics events a month, 500 PostHog AI credits (worth $5) and 3 Inbox pull requests a month, no card, one project, one-year retention. Mushi: 50 diagnoses a month, no card; self-hosting is free.