AI Nazar
ainazar.com
- Free plan
- Web
- API
- Self-hosted
What AI Nazar is
AI Nazar provides observability for large language model API calls, with a dashboard for cost, latency, token counts, and errors. It supports OpenAI, Anthropic, and Gemini through wrappers around their official clients. The dashboard shows cost by request and model, tracks latency and errors, flags spend anomalies, and compares providers. Users can inspect prompts, responses, token counts, latency, and cost, then filter requests by model, provider, or status. Its TypeScript-first SDK offers a drop-in OpenAI client wrapper with support for streaming, function calling, and tool use. AI Nazar says telemetry is reported asynchronously and adds zero latency to API calls. The SDK and dashboard are open source under the MIT license. The service is available hosted or self-hosted; self-hosted storage options are SQLite, Postgres, and Postbase. Request metadata is sent by default, while prompt and response text are sent only when content logging is enabled. The hosted free tier retains event data for 30 days on a rolling basis. Dashboard cost, latency, and usage figures are estimates, so provider billing records remain the authoritative cost reference.
Who it suits
AI Nazar may suit developers who need to inspect and compare calls across OpenAI, Anthropic, and Gemini. Its hosted and self-hosted options cater to teams with different data-storage preferences.
Good to know, and what to watch
- Tracks cost, latency, tokens, and errors.
- Supports three named model providers.
- Self-hosted data stays within the operator’s infrastructure.
- Prompts and responses are optional logging data.
- Hosted free-tier event data is retained for 30 days.
- Dashboard usage and cost figures are estimates.
- Prompt and response text require content logging to be enabled.
Plans
Every plan and priceHosted
Self-hosted
Free plan
The verdict
AI Nazar offers request-level visibility across supported LLM providers, with hosted and self-hosted deployment options. Treat dashboard cost and usage figures as estimates and consult provider billing records for authoritative costs.
What the maker publishes
- Estimate limits
- The terms say dashboard cost, latency, and usage figures are estimates and advise consulting LLM providers’ billing records for authoritative costs. source
- Features
- The dashboard includes cost per request and model, latency and error tracking, spend anomaly detection, and multi-provider comparisons. source
- Hosted retention
- The privacy policy says event data on the free hosted tier is retained for 30 days on a rolling basis. source
- Latency
- The site says telemetry is reported asynchronously and adds zero latency to API calls. source
- Open source
- AI Nazar says its SDK and dashboard are open source under the MIT license. source
- Privacy
- The privacy policy says the SDK sends request metadata by default, while prompt and response text are sent only if content logging is enabled. source
- Providers
- The site says it supports OpenAI, Anthropic, and Gemini through wrappers around their official clients. source
- Purpose
- AI Nazar provides observability for large language model API calls, including a real-time dashboard for cost, latency, tokens, and errors. source
- Request inspection
- The site says requests can be inspected with prompt, response, token counts, latency, and cost, and filtered by model, provider, and status. source
- SDK
- The TypeScript-first SDK provides a drop-in OpenAI client wrapper and says it supports streaming, function calling, and tool use. source
- Security
- The privacy policy describes TLS encryption in transit, encryption of sensitive data at rest, need-to-know production access controls, and hashed API keys. source
- Storage
- Self-hosted deployments can use SQLite, Postgres, or Postbase, and the maker says self-hosted event data stays within the operator’s infrastructure. source
- Support
- The privacy policy directs privacy inquiries to [email protected] and says it aims to respond within five business days. source
Checked against the maker's pages on 5 Oct 2026.