Live tracking · 57+ models · ModelRank Score™

Make better AI decisions,
faster.

The intelligence layer for AI model decisions. Rankings, cost analysis, API status and alerts — so you never fly blind again.

57+Models Tracked
DailyData Updates
4Benchmark Dimensions
FreeTo Get Started
Live Tracking
57
Models
82.6
Avg Score
55
Trending ↑
Top Ranked
1
Claude 4
Anthropic
92.8
2
GPT-5
OpenAI
91.2
3
Gemini Ultra
Google
90.5
Dashboard
Leaderboard
Overall Reasoning Coding Speed
1
Claude 4
92.8
+1.3
2
GPT-5
91.2
+1.4
3
Gemini Ultra
90.5
-0.7
4
DeepSeek R2
89.1
+3.1
Leaderboard
Biggest Movers
DeepSeek R2
DeepSeek
↑ +3.1
Devin 2.0
Cognition
↑ +2.5
Manus
Agent
↑ +2.1
AI News
DeepSeek R2 Tops Math Leaderboards
AI News Daily · 2026-02-20
Cursor Agent Gets Background Mode
Release · 2026-02-20
Movers & News
What's New
20
Total
10
Today
13
High Impact
HIGHFeb 20
Structured Outputs with JSON Schema
GPT-5 now guarantees valid JSON output.
HIGHFeb 20
Background Agent Mode
Cursor Agent now runs autonomously.
What's New
Biggest Movers This Week

The AI landscape shifted significantly this week. Full daily tracking available in the app.

LLM↑ +3.1
DeepSeek R2
DeepSeek
Surpassed GPT-5 in mathematical reasoning benchmarks. Now ranked #2 overall.
Agent↑ +2.5
Devin 2.0
Cognition
Major coding benchmark improvements. Now leads agent category for software engineering tasks.
Agent↑ +2.5
Manus
Manus AI
Multi-app workflow automation launched. Significant jump in autonomous task completion scores.
LLM↓ -0.7
Gemini 3 Ultra
Google
Slight drop in reasoning scores following last update. Still ranked #3 overall.

Tracking 57+ models daily. Get the full picture in the app.

Download AI Rank Free →
Best AI Models Ranked — February 2026

Updated daily. Ranked across reasoning, coding, creativity and speed. Get alerts when rankings change →

#ModelCategoryOverall7dReasoningCoding
1Claude 4AnthropicLLM92.8+1.394.191.3
2GPT-5OpenAILLM91.2+1.492.093.5
3Gemini 3 UltraGoogleLLM90.5-0.791.289.8
4DeepSeek R2DeepSeekLLM89.1+3.195.387.2
5Mistral Large 3MistralLLM87.4+0.888.186.9
6Devin 2.0CognitionAgent86.9+2.582.096.1
7ManusManus AIAgent85.3+2.583.488.2
8Llama 5MetaLLM84.70.085.183.9
9Perplexity ProPerplexityBot83.2+1.184.579.3
10AutoGPT v5AutoGPTAgent82.5+2.580.185.7
11Qwen 3AlibabaLLM81.9+1.583.280.4
12Doubao ProByteDanceLLM81.2+1.780.881.9
Showing 12 of 57 modelsDownload the app for full rankings, daily alerts and reports
See All 57 Models →

Last updated: February 20, 2026 · Rankings update daily · Get notified of changes

Everything you need to track AI
📊

Daily Leaderboard

57+ models ranked across reasoning, coding, creativity and speed. Updated every day.

💰

Cost vs. Performance

Benchmark scores vs. actual API pricing side by side. Auto-updated daily via OpenRouter API. Best value model always highlighted.

🔥

Biggest Movers

See which models climbed or fell overnight. Know before your competitors do.

Model Release Radar

Alerted within hours of any new model release. Instant benchmark comparison vs. previous version — so you know immediately if it's worth switching.

🟢

API Status Monitor

Real-time uptime and latency for every model's API. Know if GPT-5 is down before your users do. 24-hour history and instant outage alerts.

Personal Watchlist

Follow specific models and only get alerted on those. Custom alerts for rank changes, score shifts, API outages and new versions.

📋

Capabilities Table

Context window, multimodal support, function calling, fine-tuning — all models side by side. Filter and compare up to 4 models instantly.

📈

Historical Charts

90-day performance trends per model. See which models are consistently improving vs. peaking. Overlay competitors for direct comparison.

🧭

Model Recommender

Answer 3 questions about your use case and budget. Get a personalized recommendation with cost breakdown and reasoning — updated when better options emerge.

🧮

API Cost Calculator

Input your expected monthly calls and token usage. See exact monthly cost per model side by side. Export for team budget decisions.

👥

Community Benchmarks

Real-world scores submitted by developers running models locally and via API. The only crowdsourced AI benchmark leaderboard — gets more accurate as the community grows.

📰

Weekly Email Digest

Auto-generated every Monday. Top movers, new releases, best value model, API incidents — delivered to your inbox without opening the app.

One score to rule them all

Not just benchmarks. A weighted composite that answers the real question: which model should I actually use?

Reasoning
30%
Logic, analysis, complex problem solving
Coding
25%
Code generation, debugging, architecture
Cost Efficiency
20%
Performance per dollar of API cost
Speed
15%
Tokens per second, time to first token
Reliability
10%
API uptime × score consistency over 30 days
Methodology

ModelRank Score™ is a weighted composite index updated daily. Benchmark data is sourced from public evaluations (MMLU, HumanEval, MATH, HellaSwag), API pricing from OpenRouter, latency from direct measurements, and reliability from official status pages. Weights are reviewed quarterly. All data sources and update timestamps are visible in the app for full transparency.

Full methodology → modelrank.app/#methodology
Know first. Every morning.
🔥
AI Rank Daily Brief
3 models changed overnight. Claude 4 moved +1.3 — tap to see full rankings.
8:00 AM
🚨
Major ranking change
DeepSeek R2 just rose +3.1 points. Now ranked #2 overall.
Now
New model detected
Mistral Large 3 just entered the leaderboard at #8.
2m ago
Start free. Upgrade anytime.
Free
$0/mo
 
  • Top 10 models only
  • Data delayed 48 hours
  • No push alerts
  • No reports
  • Limited news feed
Download Free

Monthly plan also available at $14.99/month · Cancel anytime · Billed through App Store

⚡ Lifetime — Early Adopter
$149 once
Everything in Pro · Forever · No subscription
Get Lifetime Access →
⚙️ API Access — For Developers & Teams
Starter API
$29/mo
or $249/year · save 29%
✓ 1,000 calls/month
✓ All 7 endpoints
✓ ModelRank Score™ feed
✓ JSON + Bearer auth
✓ All Pro features
Get Starter API →
MOST POPULAR
Team API
$99/mo
or $890/year · save 25%
✓ 10,000 calls/month
✓ All 7 endpoints
✓ Priority support (24h)
✓ JSON + Bearer auth
✓ All Pro features
Get Team API →
Documentation at modelrank.app/api · Bearer token auth · Includes all Pro features
Put ModelRank Score™ on your site

Free to embed. Shows live ModelRank Score™ for any model. Updates automatically every day.

Badge Style
🏆
Claude 4 Opus
ModelRank Score™ 92.8
modelrank.app
Horizontal · Compact · Links to modelrank.app
Minimal Style
ModelRank Score™
92.8
Claude 4 Opus · #1
Vertical · Card · Auto-updates daily
Embed Code
<script src="https://modelrank.app/widget.js"></script>
<div class="mrs-widget"
     data-model="claude-4-opus"
     data-style="badge"
     data-theme="dark">
</div>

Free forever · Loads async · Links back to modelrank.app · Available to all users via the app's model detail page

We're here to help

📧 Contact Us

For questions, bug reports, or feature requests — reach out directly. We respond within 24 hours.

[email protected]

💳 Subscription Help

To manage, cancel, or get a refund for your subscription, visit your App Store or Google Play subscription settings.

Manage iOS Subscription →

Manage Android Subscription →

🔄 Restore Purchase

If you've lost access to Pro after reinstalling the app, go to Settings → Account → Restore Purchase inside the app.

Contact support if issue persists

🐛 Report a Bug

Found something that doesn't work? Please let us know with your device model, OS version, and what happened.

Send bug report →