20 tools tracked
Avg 7d gain: +25
AI Models / APIs
Access frontier AI models through unified APIs. Compare performance, manage costs, and integrate intelligence into any product.
Heat scores updated daily from 30+ live sourcesEmbed this scoreboard →
Movers this week
1
rising
rising
rising
peak
rising
stable
rising
rising
rising
rising
emerging
declining
emerging
declining
declining
declining
declining
declining
fading
stable
Claude
Teams & Solo
+24
7d
95
Heat
2
DeepSeek
Teams & Solo
+59
7d
78
Heat
3
Poe
Teams & Solo
+15
7d
72
Heat
4
Hugging Face
Teams & Solo
+2
7d
71
Heat
5
Perplexity AI
Teams & Solo
+38
7d
58
Heat
6
Mistral AI
Teams & Solo
+10
7d
56
Heat
7
ChatGPT
Teams & Solo
+21
7d
54
Heat
8
Amazon Bedrock
Teams & Solo
+28
7d
47
Heat
9
ClawBench
Teams & Solo
+25
7d
36
Heat
10
Meta Llama
Teams & Solo
+32
7d
36
Heat
11
AIGEX API
Teams & Solo
+23
7d
28
Heat
12
Anthropic API
Teams & Solo
28
Heat
13
DAAI
Teams & Solo
+24
7d
27
Heat
14
Gemini
Teams & Solo
27
Heat
15
Groq
Teams & Solo
25
Heat
16
OpenRouter
Teams & Solo
21
Heat
17
xAI Grok
Teams & Solo
20
Heat
18
Cohere
Teams & Solo
18
Heat
19
Fireworks AI
Teams & Solo
14
Heat
20
FerryAPI
Teams & Solo
5
Heat
Popular comparisons
Frequently asked questions
What is tracked under AI Models / APIs?
This category covers the foundation models and the APIs that serve them — the intelligence you call from your own product. DeepSeek, Claude, Hugging Face, and Mistral AI are tracked here, spanning proprietary frontier models, open-weight releases, and the hubs that host them.
How do I compare models for a project?
Benchmark against your own task, not a leaderboard. Score, latency, and price per token pull in different directions, and the right pick depends on whether you're doing high-volume classification or low-volume deep reasoning. Route through a layer that lets you swap models without rewriting your app, so you can chase the best cost-to-quality ratio as new releases land. For the reasoning end of the spectrum, see the distilled reasoning models guide.