Real-time pricing from OpenRouter API + token flow data from Portkey — the two largest public windows into global AI traffic.
Live data from openrouter.ai/api/v1/models
AI Gateway Traffic Comparison
Enterprise consolidates; developers diversify. Portkey handles 1.9T tokens/day for enterprise customers routing through 2-3 providers. OpenRouter serves 3,200B+ tokens/day across 400+ models — the long tail of AI demand.
Both gateways are passthrough — tokens are already counted in provider totals. This is a window into demand distribution, not additive volume.
Source: openrouter.ai/rankings · Estimated from public data
Model Share by Gateway
Why this matters: Portkey serves enterprise/production AI (24K orgs) while OpenRouter skews developer/prosumer (2.5M users). Comparing their model rankings reveals whether enterprise and developer preferences align. Key finding: Claude dominates Portkey (~30% of traffic) while OpenRouter shows more model diversity — suggesting enterprises consolidate on fewer providers while developers experiment broadly.
—
Models Available
—
Providers
—
Cheapest Output $/1M
—
Most Expensive $/1M
—
Free Models
Output Price Distribution $/1M tokens · OpenRouter live data
Models by Provider · OpenRouter catalogue
Model Price Comparison — output $/1M tokens, live
All Models
Model ↕
Provider ↕
Input $/1M ↕
Output $/1M ↕
Context ↕
Modality
Pricing pulled live from OpenRouter API. Prices shown are OpenRouter's rates (small markup over direct provider pricing).
Rankings and usage volume data are not available via public API — see openrouter.ai/rankings for usage data.