Edxperimental Labs logoEdxperimental Labs

Independent AI benchmarking from India

Independent analysis of AI

Understand the AI landscape to choose the best model, provider, and deployment strategy for your use case.

ANNOUNCEMENTEdxperimental Labs is becoming an AI benchmarking studioIndependent analysis for model selection, cost, speed, and Indian enterprise workflows.NOW BUILDINGIndian Workflow BenchmarkEvaluating LLMs on support, finance, legal, sales, document, and multilingual tasks.

Highlights

Intelligence

Edxperimental Intelligence Index; higher is better

GPT-5
62
Claude
58
Gemini
57
Grok
55
Qwen
53
Llama
52
DeepSeek
49
Mistral
44
Kimi
39
Nemo
36

Speed

Output tokens per second; higher is better

Groq
210
Together
168
Fireworks
132
OpenAI
111
Anthropic
86
Google
74
Cerebras
61
Azure
45
AWS
34
Local
28

Price

USD per 1M output tokens; lower is better

Llama
18
Qwen
28
Gemini
35
DeepSeek
48
Mistral
62
GPT
75
Claude
91
Grok
116
Nova
139
Reason
156
Get model recommendations based on your priorities for intelligence, speed, and cost.Personalized Model RecommendationCompare AI agents across capabilities, pricing, workflow reliability, and platform support.Agent Benchmark ExplorerWhat should an Indian enterprise benchmark before deploying AI?Read the evaluation methodology

Intelligence

Intelligence of leading AI models based on independent evaluations, Indian workflow tasks, and reproducible benchmark runs.

Edxperimental Intelligence Index

A prototype index combining reasoning, coding, instruction following, multilingual utility, latency, and cost efficiency.

GPT-5
62
Claude
58
Gemini
57
Grok
55
Qwen
53
Llama
52
DeepSeek
49
Mistral
44
Kimi
39
Nemo
36
Local
31
Small
27

Benchmarking

Run reproducible evaluations across LLM quality, latency, price, and Indian business workflows.

Consulting

Help teams choose providers, design eval suites, and reduce AI deployment risk before production.

Infrastructure

Publish transparent methodology, datasets, reports, and dashboards that buyers can trust.