Analytics that tell the whole story.
Events, metrics, logs, business data — one platform. The full picture.
curl -sSfL https://tell.rs | bashSingle Rust binary. Memory-safe, no garbage collector. 64M events/sec. Self-host or use Tell Cloud.
- Events — User actions, page views, sessions, revenue. Custom properties. Identity resolution across devices.
- Logs — Structured logging with 9 severity levels. Anomaly detection. First-class, not bolted on.
- Metrics — Gauges, counters, histograms from agents and SDKs. Label-based breakdown.
- Business data — Pull revenue, ad spend, traffic, and operational data from tools you already use. 15 connectors via sandboxed WASM plugins.
- Marks — Releases, deploys, incidents. Every chart knows what changed.
- SDKs — Rust (80ns/event), TypeScript, Go, Swift, Flutter, C++. Track events, identify users, structured logging.
- Sources — TCP (64M/s), Syslog, HTTP, File, Modbus.
- Connectors — Revenue (Stripe, Shopify), Marketing (Meta Ads, Klaviyo, Bitly), Development (GitHub), Infrastructure (Cloudflare), App Stores (Appbot, AppFigures, and more). Build your own with the plugin SDK.
- Analytics — Funnels, retention cohorts, lifecycle, segments, audiences, group analytics, formulas, first-time math. Breakdowns by any field. Compare periods. Aggregations: sum, avg, min, max, median, p75–p99 — per event or per user.
- Dashboards — Boards with metric cards, charts, and markdown notes. Shareable via URL. Teams with roles.
- AI — Ask questions about your data in natural language. MCP server for AI assistants. Works with any OpenAI-compatible provider. LLM-generated dashboards.
- ML — Churn prediction, anomaly detection, auto segmentation, forecasting. Per-event enrichment. (experimental)
- CLI — Full product from the terminal. Interactive TUI, live streaming, SQL queries, natural language.
- Privacy — GDPR/CCPA compliant. No cookies. PII auto-redaction. All data anonymized.
- Auth — API keys, OAuth 2.0, RBAC, workspace isolation, audit logging.
- Routing — Route logs to one place, events to another, metrics to archive — all from the same source. Content-aware rules. Fan-out to multiple destinations.
- Transforms — Pattern matching (8.1M events/sec), PII redaction (10.9M/sec), filtering, log clustering, automated IP banning.
- Signals — Real-time pub/sub on the pipeline. React to events as they happen. Build automated responses.
- Storage — ClickHouse, Parquet, Arrow-IPC, Vortex, Disk, Forwarder (Tell-to-Tell), Stdout.
| Tell | PostHog | Mixpanel | Vector | |
|---|---|---|---|---|
| Events + funnels | ✓ | ✓ | ✓ | — |
| Logs | ✓ | ✓ | — | ✓ |
| Metrics | ✓ | — | — | — |
| Business connectors | ✓ | — | — | — |
| Dashboards | ✓ | ✓ | ✓ | — |
| SDKs | ✓ | ✓ | ✓ | — |
| AI + MCP | ✓ | ✓ | ✓ | — |
| Real-time signals | ✓ | — | — | — |
| Data pipeline | ✓ | — | — | ✓ |
| Routing + transforms | ✓ | — | — | ✓ |
| Storage backends | Multiple | ClickHouse | — | ✓ |
| Self-host | ✓ | Docker | — | ✓ |
| Throughput | 64M/s | ~100K/s | — | 86 MiB/s |
| Language | Rust | Python | Go | Rust |
Apple M4 Pro · 12 cores · 5 clients · batch 500
| Source | b500 | b100 | b10 | b3 |
|---|---|---|---|---|
| TCP Binary | 64M | 55M | 6.9M | 1.4M |
| HTTP FBS | 24M | 8.3M | 1.0M | 321K |
| HTTP JSON | 2.1M | 2.3M | 848K | 307K |
| Syslog TCP | 8.7M | 8.5M | 8.3M | 2.0M |
| Sink | Events/s | Time | Written | Ratio |
|---|---|---|---|---|
| disk_binary | 33.0M | 303ms | 639 MB | 0.36x |
| disk_binary_lz4 | 21.6M | 462ms | 145 MB | 0.08x |
| disk_plaintext | 1.4M | 7.37s | 2.7 GB | 1.53x |
| disk_plaintext_lz4 | 1.1M | 9.22s | 896 MB | 0.51x |
| arrow_ipc | 2.9M | 3.49s | 1.7 GB | 0.98x |
| parquet_snappy | 2.2M | 4.58s | 249 MB | 0.14x |
| parquet_lz4 | 2.6M | 3.81s | 251 MB | 0.14x |
| parquet_zstd | 2.3M | 4.38s | 171 MB | 0.10x |
| parquet_uncompressed | 2.6M | 3.91s | 769 MB | 0.43x |
| vortex | 1.7M | 5.94s | 1.8 GB | 0.99x |
Ratio = bytes on disk / 1.8 GB input.
- tell-rs — Rust (80ns/event, 10M/s delivered)
- tell-js — TypeScript
- tell-go — Go
- tell-cpp — C++
- tell-swift — Swift
- sdk-flutter — Flutter
- witness — Host agent for logs + system metrics. 1.2 MB binary (Vector: 46 MB). 76ns/log, 30ns/metric, 4 MB idle memory. MIT.
- signals-sdk — SSE client for Tell's signal bus. Build reactive agents. MIT.
- signals-apps — Example apps on signals-sdk. Reference: ip-ban agent.
Built by the founder of Logpoint (European SIEM, acquired).