AI Model Rankings

Rankings and statistics based on real usage data across Onysoft AI Gateway.

87M
Weekly Tokens
70K
Weekly Usage
690
Active Models
70K
Total Requests
📊 Most Used Models
Weekly model usage (by tokens, last 12 weeks)
1
openai/gpt-4o
18.0M tokens
12,000 uses
2
anthropic/claude-sonnet-4.6
14.0M tokens
9,500 uses
3
google/gemini-1.5-flash
11.0M tokens
8,800 uses
4
openai/gpt-4-turbo
9.0M tokens
6,200 uses
5
anthropic/claude-haiku-4.5
8.0M tokens
11,000 uses
6
openai/gpt-3.5-turbo
7.0M tokens
9,200 uses
7
meta-llama/llama-3.1-70b
6.0M tokens
4,100 uses
8
mistralai/mixtral-8x7b
5.0M tokens
3,600 uses
9
google/gemini-1.5-pro
5.0M tokens
2,900 uses
10
deepseek/deepseek-v3
4.0M tokens
2,700 uses
🏆 Leaderboard
This week's most popular models (by token usage)
1
openai/gpt-4o
openai
18.0M tokens
12,000 uses
2
anthropic/claude-sonnet-4.6
anthropic
14.0M tokens
9,500 uses
3
google/gemini-1.5-flash
google
11.0M tokens
8,800 uses
4
openai/gpt-4-turbo
openai
9.0M tokens
6,200 uses
5
anthropic/claude-haiku-4.5
anthropic
8.0M tokens
11,000 uses
6
openai/gpt-3.5-turbo
openai
7.0M tokens
9,200 uses
7
meta-llama/llama-3.1-70b
meta-llama
6.0M tokens
4,100 uses
8
mistralai/mixtral-8x7b
mistralai
5.0M tokens
3,600 uses
9
google/gemini-1.5-pro
google
5.0M tokens
2,900 uses
10
deepseek/deepseek-v3
deepseek
4.0M tokens
2,700 uses
💰 Cost Analysis
Average cost comparison per model (last 30 days, USD / 1M tokens)
1
google/gemini-1.5-flash
google
$0.35 / 1M
11.0M tokens
2
deepseek/deepseek-v3
deepseek
$0.45 / 1M
4.0M tokens
3
mistralai/mixtral-8x7b
mistralai
$0.65 / 1M
5.0M tokens
4
meta-llama/llama-3.1-70b
meta-llama
$0.88 / 1M
6.0M tokens
5
anthropic/claude-haiku-4.5
anthropic
$1.25 / 1M
8.0M tokens
6
openai/gpt-3.5-turbo
openai
$1.50 / 1M
7.0M tokens
7
google/gemini-1.5-pro
google
$5.25 / 1M
5.0M tokens
8
openai/gpt-4o
openai
$7.50 / 1M
18.0M tokens
9
anthropic/claude-sonnet-4.6
anthropic
$9.00 / 1M
14.0M tokens
10
openai/gpt-4-turbo
openai
$15.00 / 1M
9.0M tokens
📅 Daily Usage Trend
Request and token trend over the last 30 days
⚡ Speed Comparison
Average response speed (tokens/second, last 7 days)
1
google/gemini-1.5-flash
google
180 tok/s
320ms avg.
2
anthropic/claude-haiku-4.5
anthropic
150 tok/s
380ms avg.
3
openai/gpt-3.5-turbo
openai
130 tok/s
420ms avg.
4
deepseek/deepseek-v3
deepseek
95 tok/s
680ms avg.
5
mistralai/mixtral-8x7b
mistralai
90 tok/s
720ms avg.
6
meta-llama/llama-3.1-70b
meta-llama
75 tok/s
850ms avg.
7
openai/gpt-4o
openai
68 tok/s
950ms avg.
8
google/gemini-1.5-pro
google
60 tok/s
1,050ms avg.
9
openai/gpt-4-turbo
openai
45 tok/s
1,100ms avg.
10
anthropic/claude-sonnet-4.6
anthropic
42 tok/s
1,200ms avg.
Want help finding the right model?