# Meilisearch Problems & Issues 2026 - Cloud Outages Exposed

Source: https://www.tryorbye.com/products/meilisearch
Last analyzed: 2026-03-05
Try or Bye score: 72/100

## Verdict: Try

Meilisearch is worth trying for most users (score 72/100). Most reported issues: Meilisearch Cloud suffers frequent outages; Service jams with large datasets (30M+ docs); Indexing falls behind by 30-50 minutes on updates. Not a good fit for: Companies needing enterprise-grade reliability; Teams with millions of frequently-updated records; Companies requiring per-user permissions. Users most often switch to OpenSearch, Apache Solr, Algolia.

## What it is

Meilisearch is an open-source, lightning-fast search engine written in Rust. Available self-hosted (free) or via Meilisearch Cloud. Features instant search, typo tolerance, hybrid search with AI, and faceted filtering. Simple setup but struggles at scale with millions of documents and frequent updates.

## Common problems

- **Meilisearch Cloud suffers frequent outages** (reliability): Multiple G2 reviewers report cloud service reliability issues. One user describes Meilisearch Cloud as 'just a bunch of incompetent people unable to solve any issues' with 'search freezes lasting for three hours frequently.' Cloud resiliency needs significant improvement.
- **Service jams with large datasets (30M+ docs)** (performance): GitHub issues report that after importing 30 million documents, 'the service experiences jams three or four times daily, each lasting several to ten minutes, making the service unavailable.' Large datasets require careful management.
- **Indexing falls behind by 30-50 minutes on updates** (performance): For frequently changing data, Meilisearch struggles to keep up. Users report 'updating the index on single events like user starring a post is not effective' with 'the index getting backlogged and behind by 30-50 minutes.' Not suited for real-time update scenarios.
- **Documents silently fail to index without warnings** (reliability): GitHub issue #4438 documents how 'when indexing new documents through the JS API, Meilisearch returns a JSON indicating a successfully enqueued task, but documents may silently fail to index without logging or meaningful response.' No error visibility.
- **Cloud pricing expensive at high traffic** (pricing): G2 reviews note 'if you have a lot of traffic, the pricing per search on the cloud can be expensive.' Recent pricing changes have raised concerns, as 'the new pricing model might not feel as fair or cost-effective for those with large indexes but lower usage.'
- **Only searches first 100k of 1M+ records** (performance): Users report that 'despite successfully importing 1 million+ records, Meilisearch doesn't search 9/10ths of the records, searching maybe only the first 100k.' This creates a critical gap where most data is effectively invisible.
- **Dashboard lacks sophistication for index management** (usability): Users feel 'the dashboard should be more sophisticated to effectively manage indexes.' The dashboard doesn't provide insights on search requests like response times and top queries. Advanced use cases require workarounds or external tooling.
- **Storage bloat: 10MB data becomes 200MB indexed** (performance): Hacker News users raise concerns about storage efficiency: 'a 10MB document is taking approximately 200MB according to their documentation.' This 20x storage multiplier becomes expensive at scale and limits dataset size.
- **Missing auth/permissions support requires workarounds** (usability): Users report having to 'use workarounds to handle lacking auth/permissions support.' Meilisearch doesn't natively support per-user search permissions, requiring Redis or other tools to manage access control.
- **Sort order breaks after 4-5 pages of results** (reliability): GitHub issue #5180 documents that 'sort order is not working as expected, with results resetting after 4-5 pages.' Pagination with sorting becomes unreliable for users browsing deep into result sets.

## What users like

- **Lightning-fast search with sub-50ms response times**: Meilisearch delivers exceptionally fast search performance. Users consistently praise 'lightning-fast and highly relevant search results' with typical response times under 50ms. Built in Rust for maximum efficiency.
- **Extremely easy setup and developer-friendly**: Developers love the simplicity: 'seamless integration with Laravel, ease of use, nice cloud dashboard.' Can be set up in minutes with excellent SDKs for all major languages. Documentation is comprehensive and approachable.
- **Self-hosted is completely free and MIT licensed**: Unlike Algolia or Typesense's GPL license, Meilisearch's MIT license provides maximum flexibility. Self-hosted version is 100% free with no limitations. No licensing fees or hidden costs for self-hosting.
- **Typo tolerance and instant search built-in**: Out-of-the-box typo tolerance, synonyms, and instant-as-you-type search work without configuration. Users don't need to worry about misspellings. Makes search feel polished from day one.
- **Excellent client libraries for all languages**: Strong SDKs for JavaScript, Python, PHP, Ruby, Go, Java, Rust, and more. First-party integrations with Laravel, Rails, and other frameworks. Official Docker images for easy deployment.
- **Hybrid search with AI-powered semantic capabilities**: Recent releases added hybrid search combining full-text and AI-powered semantic search. Supports vector embeddings from OpenAI, Cohere, or Hugging Face models. Balances keyword matching with contextual understanding.

## Pricing

- **Self-Hosted (Open Source)**: $0 (free) — hidden costs: You pay for infrastructure (servers, hosting, maintenance). No official support included.
- **Build (Cloud)**: $30 (monthly) — hidden costs: Overages charged for additional searches. Pricing scales with document count and search volume.
- **Pro (Cloud)**: $300 (monthly) — hidden costs: AI/semantic search features may require additional compute resources. High-traffic apps exceed included searches.
- **Resource-Based (Cloud)**: usage_based — hidden costs: Larger instances (XL, 2XL, 4XL) become expensive. Contact sales for custom pricing on large deployments.

## Who should and shouldn't use it

- **Small to medium websites** — Good fit: Perfect for documentation sites, blogs, and e-commerce stores with under 1 million products. Fast setup, excellent search quality, and affordable pricing. The sweet spot for Meilisearch.
- **Developers who want simplicity** — Good fit: Meilisearch shines for developers who want Algolia-quality search without the complexity. Great SDKs, easy configuration, and instant results make it a joy to implement.
- **Budget-conscious startups** — Good fit: Self-hosted version is 100% free with MIT license. Can run on modest infrastructure and grow with your product. Cloud pricing is more predictable than Algolia.
- **Companies needing enterprise-grade reliability** — Avoid: Meilisearch Cloud has reported frequent outages and 'search freezes lasting three hours.' For mission-critical search, consider Algolia or Elasticsearch with proper SLAs.
- **Teams with millions of frequently-updated records** — Avoid: Indexing falls 30-50 minutes behind with frequent updates. Service jams occur with 30M+ documents. Better suited for static or slowly-changing datasets like documentation.
- **Engineering teams seeking DevOps simplicity** — Good fit: Single binary deployment, no JVM tuning, no cluster management. Rust-based efficiency means low resource usage. Much simpler to operate than Elasticsearch.
- **Sales and marketing teams (internal search)** — Depends: Works well for product catalogs and internal search. Lacks built-in analytics for conversion tracking. May need additional tooling for marketing use cases.
- **Companies requiring per-user permissions** — Avoid: No native auth/permissions support. Requires workarounds with Redis or custom middleware. Elasticsearch or Algolia handle multi-tenant permissions better out of the box.

## Alternatives

- **OpenSearch**: Companies switch for Elasticsearch-like power without licensing concerns. Gain: 100% open source, AWS-backed development, Elasticsearch compatibility. Trade-off: Same operational complexity as Elasticsearch.
- **Apache Solr**: Enterprises switch for mature, battle-tested search with no vendor lock-in. Gain: decades of production use, extensive plugin ecosystem, free forever. Trade-off: Complex setup, less developer-friendly than modern alternatives.
- **Algolia** (https://www.tryorbye.com/products/algolia): Companies switch for enterprise reliability and global CDN. Gain: 99.999% uptime SLA, world-class support, comprehensive analytics and A/B testing. Trade-off: Expensive at scale, pricing unpredictable with traffic spikes.
- **Elasticsearch** (https://www.tryorbye.com/products/elasticsearch): Teams switch for true distributed architecture and scale. Gain: horizontal scaling across nodes, battle-tested at petabyte scale, rich analytics. Trade-off: Operational complexity, JVM tuning required, steep learning curve.
- **Typesense** (https://www.tryorbye.com/products/typesense): Teams switch for better large dataset handling and replicated cluster architecture. Gain: high availability via Raft consensus, dynamic query-time parameters without reindexing, fixed cluster pricing. Trade-off: GPL license vs MIT, smaller ecosystem.

## FAQ

### Is Meilisearch worth it in 2026?

Meilisearch is excellent for small to medium sites needing fast, easy-to-implement search. Self-hosted is free (MIT license) and performs well under 1M documents. However, avoid Meilisearch Cloud for mission-critical apps due to reliability concerns. For large datasets or frequent updates, consider Typesense or Elasticsearch instead.

### What are the main Meilisearch problems?

The most reported Meilisearch issues are: 1) Cloud service instability with multi-hour outages, 2) Poor performance with large datasets (30M+ documents cause service jams), 3) Indexing falls behind 30-50 minutes with frequent updates, 4) Storage bloat (10MB data becomes 200MB indexed), and 5) Silent indexing failures without error messages.

### Is Meilisearch better than Algolia?

Meilisearch is better for budget-conscious teams and developers who want simplicity. It's free to self-host and has lower cloud pricing. However, Algolia offers superior reliability (99.999% uptime), better enterprise features, and world-class support. Choose Meilisearch for small-medium sites, Algolia for mission-critical enterprise search.

### Can Meilisearch handle millions of documents?

Technically yes, but with significant caveats. Users report indexing slows dramatically at scale, with one customer having 311M documents. However, frequent updates cause 30-50 minute backlogs, and some users found only 100k of 1M+ records were searchable. For millions of frequently-updated documents, Elasticsearch or Typesense are safer choices.

### Is Meilisearch Cloud reliable?

Meilisearch Cloud reliability is a common complaint. G2 reviewers report 'search freezes lasting three hours frequently' and describe the cloud service as having recurring issues. The official status page shows minimal downtime, but user reviews paint a different picture. For production critical apps, self-host or consider alternatives.

### What is the best Meilisearch alternative?

Typesense is the closest alternative with similar simplicity but better large-dataset handling. Algolia offers superior enterprise features at higher cost. OpenSearch provides Elasticsearch-like power for free. For e-commerce with high uptime needs, Algolia. For budget-conscious teams, Typesense. For massive scale, Elasticsearch/OpenSearch.

### How much does Meilisearch cost?

Self-hosted Meilisearch is free forever (MIT license). Meilisearch Cloud starts at $30/month (Build plan with 50k searches). Pro plan is $300/month with 250k searches. Resource-based pricing lets you pick CPU/RAM for predictable costs. Overages apply when exceeding search quotas. Much cheaper than Algolia at similar scale.

### Why is Meilisearch indexing so slow?

Meilisearch indexing slows because it prioritizes search speed over index speed. Memory-intensive indexing on large batches uses significant RAM. Tips: send larger batches (150MB max), configure settings before adding documents (reindexing is slow), and use SSDs. For frequently-changing data, consider Typesense which handles updates better.

### Does Meilisearch support AI search?

Yes, as of 2025 Meilisearch supports hybrid search combining full-text and AI semantic search. You can use vector embeddings from OpenAI, Cohere, or Hugging Face. Adjust the semantic ratio to balance keyword vs meaning-based results. Note: AI features may require more resources and are newer/less battle-tested than core search.

### Can I self-host Meilisearch for free?

Yes, Meilisearch is fully open source with MIT license. Unlike competitors with restrictive licenses, you can self-host without any cost or limitations. Single binary deployment makes it simple. You only pay for your own infrastructure. No hidden fees, no enterprise-only features locked away.

### Is Meilisearch good for e-commerce?

Meilisearch works well for small to medium e-commerce sites (under 100k products). Fast search, typo tolerance, and faceted filtering work great. However, for large catalogs with frequent inventory updates, indexing lag becomes problematic. Enterprise e-commerce should consider Algolia for reliability or Elasticsearch for scale.

---
Try or Bye aggregates problem patterns from user reviews, forums, and app stores. Methodology: https://www.tryorbye.com/methodology