Headline benchmark standings and comparison context.
Top benchmark results for qwen/qwen2.5-72b.
Key dates, capabilities, and model metadata.
Parameters
7,610,000,000
Training Tokens
18,000,000,000,000
Input
No modalities listed.
Output
No modalities listed.
Detailed benchmark comparisons now live in the Compare tool.
Based on the last hour of usage: we total provider spend, divide by total token volume, and express it as USD per 1M tokens. Providers with more traffic naturally carry more weight.
Weighted Avg Input Price
--
per 1M tokens (past hour)
Weighted Avg Output Price
--
per 1M tokens (past hour)
| Provider | Input $/Million | Output $/Million | Cache Token % |
|---|---|---|---|
Alibaba Cloud | $1.400 | $5.600 | -- |
DeepInfra | -- | -- | -- |
NovitaAI | $0.380 | $0.400 | -- |
Together | -- | -- | -- |
Meter
Public apps observed in gateway request traffic for this model.
Core latency and throughput trends from recent traffic.
Latency
Throughput
Uptime
Total Context
Max Output
Latency
Throughput
Uptime
Total Context
Max Output
Latency
Throughput
Uptime
Total Context
Max Output
Latency
Throughput
Uptime
Start calling this model with endpoint-specific examples.
# 1) Set your key
export AI_STATS_API_KEY="aistats_***"
# 2) Send a request
curl -s https://api.phaseo.app/v1/responses \
-H "Authorization: Bearer $AI_STATS_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen/qwen2.5-72b",
"input": "Give me one fun fact about cURL."
}'