Skip to content
SPECIFICTION
Get listed

Freezetime

benginn.github.io

From €4.99/mo Free plan Open source WebmacOSLinuxAPISelf-hosted
Visit benginn.github.io
Spec score 9.1 #2 of 22 in Game Replay Analyzers

What it is

Freezetime is a free, self-hosted analysis platform for Counter-Strike 2 demos. Its public archive is a read-only site with more than 2,600 professional matches, searchable and grouped by series, and needs no installation or account. For a personal demo, the Analyze page parses one file in the browser using WebAssembly, so the file stays on the user's machine. Processing produces 16 Hz player positions along with kills, grenades, and economy data. Synchronized 2D replays show player health, economy, kill feed, grenade trajectories, smoke, and flash effects. Analysis tools include opponent reports, predictions, ML Lab, leaderboards, match comparison, heatmaps, ghost rounds, Moments, and Pattern Finder. My DB can process private demos without the server keeping a copy. Self-hosting unlocks features marked self-host only, but users must supply their own demos because the repository contains no match data or database dumps. The project says smart features use deterministic methods and locally trained models rather than external AI services. Freezetime is MIT licensed, with no warranty.

Spec sheet

What it does
Freezetime analyzes Counter-Strike 2 professional and personal demo files for replay and tactical reporting.Read 15 Oct 2026 · csfreezetime.com
Purpose
Freezetime is a self-hosted analysis platform for Counter-Strike 2 demos.Read 1 Oct 2026 · github.com
Support
The project directs questions and setup help to GitHub Discussions and bugs or feature requests to GitHub Issues.Read 1 Oct 2026 · github.com
License
The project is released under the MIT license, permitting use, modification and sharing.Read 15 Oct 2026 · github.com
2D replay
Every archived round can be scrubbed, followed by player, jumped to kills, or stacked as ghost rounds.Read 15 Oct 2026 · csfreezetime.com
Account data
A signed-in account stores Steam identity details, plan, a 30-day session cookie and optional notes, bookmarks, playlists and searches.Read 15 Oct 2026 · csfreezetime.com
Analysis
It includes opponent reports, predictions, ML Lab, leaderboards, match comparison, heatmaps, ghost rounds, Moments and Pattern Finder.Read 1 Oct 2026 · github.com
API service
The self-hosted stats service serves the API and web application on port 8090.Read 15 Oct 2026 · github.com
Archive coverage
The archive collects S-, A- and B-tier events and qualifiers from HLTV, while independent C-tier events are excluded.Read 15 Oct 2026 · csfreezetime.com
Archive freshness
New demos are collected weekly and published overnight rather than immediately after each match.Read 15 Oct 2026 · csfreezetime.com
Archive size
The public archive lists 2,985 professional maps across 48 tournaments and 107 teams.Read 15 Oct 2026 · csfreezetime.com
Browser parsing
Personal .dem files are parsed in the browser with WebAssembly and are not uploaded to Freezetime servers.Read 15 Oct 2026 · csfreezetime.com
Data retention
Archived matches are kept indefinitely, while older seasons receive less weight in tendencies and predictions using a 90-day half-life.Read 15 Oct 2026 · csfreezetime.com
Demo processing
Each demo is parsed into 16 Hz positions, kills, grenades and economy, with hundreds of features precomputed in database tables.Read 1 Oct 2026 · github.com
Documentation depth
Documentation includes a plain-language architecture guide, a Turkish design document, service locations and Docker-based quickstart instructions.Read 15 Oct 2026 · github.com
Independent description
Trendshift categorizes the repository as a self-hosted coach-grade CS2 analyzer with locally trained predictions and no external AI.Read 15 Oct 2026 · trendshift.io
Independent launch scope
The author’s public-site announcement describes a zero-setup free site with more than 2,600 professional matches and weekly updates.Read 15 Oct 2026 · reddit.com
Independent repository metadata
Trendshift lists the project as TypeScript with one contributor and an MIT license.Read 15 Oct 2026 · trendshift.io
Independent self-host scope
The announcement says self-hosting adds archive-wide grenade search, scenario queries, veto simulation, playlists and coach notes beyond the static site.Read 15 Oct 2026 · reddit.com
Limit: bring your own demos
Self-hosted users must supply GOTV, FACEIT or other demo files to build their archive.Read 15 Oct 2026 · reddit.com
Limit: setup effort
The author warns that the self-hosted version requires Docker and several services, so it is not a click-and-go website.Read 15 Oct 2026 · reddit.com
Local operation
Freezetime runs on the user's own machine and says its smart features use deterministic statistics, geometry and locally trained models without external AI services.Read 1 Oct 2026 · github.com
Local processing
The product says its statistical, geometric and model features run locally without external AI services or per-use charges.Read 15 Oct 2026 · csfreezetime.com
Not a Valve product
The official footer states that demo files and radar images belong to Valve and tournament organizers and that Freezetime is not affiliated with Valve.Read 15 Oct 2026 · csfreezetime.com
Opponent reports
Reports cover utility habits, early setups, executes, rotations, roles and economy for each team and map.Read 15 Oct 2026 · csfreezetime.com
Own demos
The Analyze page parses a single demo in the browser using WebAssembly, and the file never leaves the user's machine.Read 1 Oct 2026 · github.com
Pattern Finder
Pattern Finder maps grenade landing locations and timing and lets users isolate areas by dragging a map box.Read 15 Oct 2026 · csfreezetime.com
Player progress
Personal demos can generate rating history, role metrics and suggestions compared with a role median.Read 15 Oct 2026 · csfreezetime.com
Praise: feature breadth
The project announcement highlights synchronized replay, opponent reports, grenade pattern search, scenario queries and tested predictions as its core strengths.Read 15 Oct 2026 · reddit.com
Praise: privacy
A Reddit commenter called the in-browser parser impressive, while the project author says demos never leave the machine.Read 15 Oct 2026 · reddit.com
Predictions
Win probabilities and opponent predictions are trained on the archive and shown with a model-performance laboratory.Read 15 Oct 2026 · csfreezetime.com
Private database
My DB lets users process private demos without touching the main archive, and the server does not keep a copy.Read 1 Oct 2026 · github.com
Public archive
The public GitHub Pages site is a free, read-only archive that requires no installation or account.Read 1 Oct 2026 · github.com
Replay
The platform provides synchronized 2D replays with player health, economy, kill feed, grenade trajectories, smoke and flash effects.Read 1 Oct 2026 · github.com
Round search
Users can filter rounds by team, map, economy, result and in-round events such as AWP kills or late defuses.Read 15 Oct 2026 · csfreezetime.com
Self-host architecture
The documented stack uses Rust parsing, Python enrichment and machine learning, Go services, React/PixiJS, PostgreSQL, ClickHouse, NATS JetStream and MinIO under Docker Compose.Read 15 Oct 2026 · github.com
Self-host data requirement
The repository supplies the analysis engine but not match data or database dumps, so operators must provide their own demos.Read 15 Oct 2026 · github.com
Self-host limits
Presence queries, Scenarios, free-range heatmap windows and cross-match round overlays are marked self-host only on the public site.Read 1 Oct 2026 · github.com
Self-hosting
Self-hosting requires supplying your own demos because the repository ships the engine but no match data or database dumps.Read 1 Oct 2026 · github.com
Static-site limits
The architecture notes that presence queries, Scenario Lab, free-range heatmap windows and cross-match round overlays require the live self-hosted backend.Read 15 Oct 2026 · github.com
Team workspace
Team editions are described as shared workspaces with playlists, notes, playbooks and roster progress boards.Read 15 Oct 2026 · csfreezetime.com
Technical stack
The system uses PostgreSQL, ClickHouse, MinIO and NATS JetStream in Docker, with Rust, Go, Python and React components.Read 1 Oct 2026 · github.com

Plans and prices

Free Free public archive · self-hosted full studio · no per-use cost Read 1 Oct 2026 · github.com
Go €4.99 a month Full studio during beta; payments open at launch Read 15 Oct 2026 · csfreezetime.com
Pro €29.99 a month Full studio during beta; payments open at launch Read 15 Oct 2026 · csfreezetime.com
Max €49.99 a month Full studio during beta; payments open at launch Read 15 Oct 2026 · csfreezetime.com

Every Freezetime plan and price →

Strong points and limits

Strong points

  • Public archive contains more than 2,600 professional matches.
  • Analyze parses a demo in-browser using WebAssembly.
  • Replays show health, economy, kill feed, and grenade effects.
  • My DB handles private demos without retaining a server copy.
  • MIT licensed and free.

Limits

  • The public archive is read-only.
  • Self-hosting requires users to supply demos.
  • The repository has no match data or database dumps.
  • The license provides no warranty.

Who it suits

Freezetime suits Counter-Strike 2 players or analysts who want demo replays and match analysis. The public archive needs no setup; users wanting the full self-hosted studio must provide their own demos.

Freezetime in depth

Read the full write-up

Freezetime is a Counter-Strike 2 analysis platform for players who want to study professional matches or inspect their own demos. Its strongest case is the combination of a substantial public archive and a capable self-hosted studio; the latter, however, depends on supplying your own match data.

Overview

The public GitHub Pages archive offers more than 2,600 professional matches, grouped by series and searchable. It is free, read-only and needs neither an account nor installation, making it a low-friction way to explore existing games. The archive is not a place to upload personal matches: the separate Analyze page parses one demo in the browser using WebAssembly, and the file stays on your machine.

Freezetime converts demos into 16 Hz player-position data, kills, grenades and economy, with hundreds of features precomputed in database tables. That data-led approach supports detailed review beyond simply replaying a round. Self-hosting requires more work: the repository provides the engine, but no match data or database dumps, so users must supply their own demos.

Key features

Replay and analysis

Synchronized 2D replays bring together player health, economy, the kill feed, grenade trajectories, smoke and flash effects. This gives round review useful tactical context, while remaining a 2D presentation rather than a conventional game replay. Analysis tools include opponent reports, predictions, ML Lab, leaderboards, match comparison, heatmaps, ghost rounds, Moments and Pattern Finder. Their breadth will suit players who want to inspect patterns and compare performance, though the full data workflow is more involved than browsing the public archive.

Private demos and local operation

My DB lets users process private demos without changing the main archive, and the server does not keep a copy. Freezetime says its smart features use deterministic statistics, geometry and locally trained models without external AI services. That is a good fit for users who want analysis on their own machine and control over private match files; it is not a hosted private-match library that removes setup and data-management work.

Some capabilities—Presence queries, Scenarios, free-range heatmap windows and cross-match round overlays—are marked self-host only on the public site. Those are reasons to run the studio locally, but they also make self-hosting necessary for users who need those specific tools.

Pricing

Freezetime is free: the Free plan costs nothing, with a public archive, a self-hosted full studio and no per-use cost. The public archive is the simplest option for readers who only need to browse professional matches. The Analyze page offers a way to parse a single personal demo without setting up the studio, but it does not replace the self-hosted environment for its full analysis workflow.

The free plan carries no stated per-use charge, but self-hosting has a practical cost in effort: users must provide demos and database data and operate the system themselves. Freezetime is MIT licensed for use, modification and sharing, with no warranty. The project directs setup questions to GitHub Discussions and bug reports or feature requests to GitHub Issues.

Platforms

Freezetime supports Linux, macOS, web and self-hosted operation, and covers Counter-Strike 2. The public archive and browser-based demo parser serve web users; the full studio is for users prepared to run the system on their own machine. Its stack uses PostgreSQL, ClickHouse, MinIO and NATS JetStream in Docker, alongside Rust, Go, Python and React components, so this is a substantial self-hosted system rather than a lightweight desktop viewer.

Who it's for

Choose Freezetime if you want a free way to study professional CS2 matches, or if you are comfortable providing match data and operating a self-hosted analysis environment. It is less suitable if you want a private demo archive that is ready to use without setup, or analysis for games beyond Counter-Strike 2.

Pros and cons

Pros: The searchable archive of more than 2,600 professional matches is free and requires no account or installation.

Pros: Browser parsing keeps a single demo on the user's machine, while My DB avoids retaining a server copy of private demos.

Pros: Synchronized replays and a wide set of reports and pattern tools give serious match review more depth than replay alone.

Cons: The public archive is read-only, so personal match analysis requires the browser parser or a separate self-hosted setup.

Cons: Self-hosting requires users to supply match data and database dumps, and the underlying system has a multi-service stack.

Cons: Several advanced analysis features are self-host only, making the ready-to-browse archive insufficient for those workflows.

Alternatives

Game Replay Analyzers is the broader category for readers comparing replay-focused tools.

DataCoach is a better fit for players who want coaching output and training plans, with its free monthly plan including 10 replays, 10 chat tokens and progression tracking.

ReplayLabs suits readers looking for a web and API service with metered upload or request plans rather than Freezetime's self-hosted studio.

SC2 Tools is another free option for players who want its features across Linux, macOS, web and Windows; voluntary tips do not unlock extra features.

CS2 Demo Analysis Kit is a free macOS and Windows alternative with an AGPL-3.0-only Studio and MIT-licensed ecosystem packages and Python exporter.

ReplayDojo is a free browser-based replay-analysis option.

CaptureAge is a Windows alternative for readers focused on its supported game, with a free tier and a paid Pro plan.

CS Demo Analyzer is a free open-source CLI for analyzing and exporting CS2 and CS:GO demos, making it a better fit for command-line workflows.

Gamescan is another web option with paid plans that include monthly credits, game modes and clip analysis.

Verdict

Freezetime is an excellent choice for CS2 players who want a free professional-match archive or are equipped to run their own demo-analysis studio. Its combination of synchronized replay, detailed statistical tools and locally handled private demos is compelling. Look elsewhere if you need a turnkey hosted workspace for personal matches or analysis beyond Counter-Strike 2.

Verdict

Freezetime offers an accessible read-only archive and browser-based parsing alongside self-hosted analysis tools. Choose the self-hosted route only if you can supply the match data it requires.

Where Specifiction places it

Alternatives to FreezetimeAll alternatives →