Skip to main content
tif1 logo tif1 logo

Welcome to tif1

tif1 is a Python library for Formula 1 data analysis. It provides access to timing, telemetry, and weather data from TracingInsights (2018-current), and to Ergast results, lap times, pit stops and championship standings (1950-current). tif1 is an alternative to fastf1 with a compatible API. It loads data lazily, fetches multi-file loads in parallel over HTTP/2, and supports pandas and polars backends.

Why tif1?

Parallel Loading

Multi-file loads run as parallel requests over HTTP/2. The benchmark suite checks that a parallel batch finishes more than 3x faster than the same requests in sequence.

API Compatible

The API follows fastf1 conventions. Most scripts migrate with few code changes.

Flexible Backends

Choose Pandas for compatibility, or Polars to get polars.DataFrame laps and telemetry.

Multi-tier Caching

Two cache tiers: an in-memory LRU cache and a SQLite cache for persistent storage.

Why tif1 over fastf1?

tif1 does not replace fastf1. The timing and telemetry data in tif1 is generated with fastf1. tif1 differs in how that data is delivered:
  1. Download only the necessary data. session.load() is not necessary. One lap of telemetry for one driver needs three small files: the driver list, the lap table, and the telemetry of that lap.
  2. No F1TV account. Since fastf1 3.9.0, most fastf1 features need an F1TV Access subscription and a sign-in. tif1 reads published files and needs no account.
  3. No rate limits on the session data. fastf1 raises RateLimitExceededError after 500 uncached requests in one hour across all its APIs, 200 of them to jolpica-f1. tif1 reads static CDN files, the Ergast tables included, and has no request quota. tif1 calls jolpica-f1 only for data that is not in the files, within the jolpica-f1 and fastf1 limits.
  4. Built-in charts. 22 optional one-call plot_*() functions turn raw data into finished charts: tif1.plot_top_speeds(...), tif1.plot_driver_laptimes(...), and more.
  5. Public CDN access. tif1 needs no API key, cookie, or account. Any network that can reach jsDelivr, Hugging Face, or StaticDelivr can load the data.
  6. More data. Mini-sector color codes on every lap row and acceleration channels in telemetry. The fastf1 laps and telemetry do not include these columns.

Full comparison

Read the “Why tif1 instead of fastf1?” page. It gives the fastf1 sign-in requirement and rate limits, with sources.

Key Features

  • Multi-CDN: jsDelivr, Hugging Face buckets, and StaticDelivr with automatic fallback and SQLite caching
  • Complete: Lap times, sectors, telemetry, tire compounds, weather, and race control messages
  • Historical: Timing and telemetry from 2018-current; Ergast results, lap times and standings from 1950-current
  • Reliable: Automatic retry logic with circuit breaker and multi-CDN fallback
  • Async: Parallel data fetching for multi-file loads
  • HTTP/2: Uses niquests for multiplexed, connection-efficient network requests
  • Flexible: Supports both pandas and polars backends
  • Type-Safe: Type hints on the public API for IDE support
  • Jupyter-Ready: tif1.jupyter adds HTML display for sessions, drivers, and laps in notebooks
  • Native Charts: 22 one-call chart functions in tif1.charts for track maps, telemetry comparisons, and performance analysis

Performance

tif1 does not publish a timed comparison with fastf1. The benchmark suite in the repository measures tif1 itself. One test sends 10 simulated requests of 0.1 s each. The parallel batch must finish more than 3x faster than the same requests in sequence. Results vary by machine and network. Run uv run pytest tests/benchmarks/ -m benchmark in a clone of the repository to measure them.

Core Architecture

tif1 uses an HTTP/2 networking stack with multi-CDN support and a multi-tier cache. This design decreases the time spent waiting for data.

Data Available

  • Lap times and sectors (S1, S2, S3). LapTime is a Timedelta on pandas and a Duration on polars. LapTimeSeconds is float seconds on both.
  • Tire compounds and stint information
  • Telemetry: speed, throttle, brake, RPM, gear, and DRS as a 0/1 flag
  • Position data (X, Y, Z coordinates)
  • Acceleration data (X, Y, Z axes)
  • Each lap’s weather sample in one call (laps.get_weather_data())
  • Driver metadata with team colors and headshots
  • Race control messages

Next Steps

Why tif1?

Why choose tif1 instead of fastf1?

Installation

Install tif1 and verify the installation.

Quickstart

Load a first session and read laps and telemetry.

Examples

Common patterns and code examples.

Charts

Visualize F1 data with 22 one-call chart functions.

Data Reference

Every field in the raw TracingInsights data files, with explanations.

Migration Guide

Coming from fastf1? See the differences.

API Reference

Explore the complete tif1 API.

Tutorials

Learn with detailed analysis examples.

Installation

Install tif1

Why tif1?

Why tif1 instead of fastf1?

Quickstart

Load a first session

API Reference

Explore the API

Charts API

Native chart functions

Tutorials

Learn with examples
Last modified on October 30, 2026