Models and cost
Many models, every cost traced
Use models from several providers, and trace every charge to a workspace, a member and an app.

This month's cost chart: daily model and tool spend over 30 days, stacked by four workspaces; $1,720 so far against a $2,000 limit.
Set up models once, with a default for each job
- 01
Model definitions
Chat, image, video, audio, embedding and rerank
- 02
API platforms
Ready on Evose Cloud; self-hosted, connect your own platforms and endpoints
- 03
Deployments
Several routes, by weighted round-robin or priority failover
- 04
Default models
Set per job: titles, summaries, embeddings, reranking, image understanding and more
One model, several routes
Requests are split by weight across providers, and fail over by priority when a route has trouble.

Every cost,traced to its source
Spending limits
Daily, weekly or monthly limits for employees, API keys and IM bots.
Drill into any cost, export reports by day
8 dimensions
Workspace, model, marketplace tool, app, department, member, API key and IM bot, each one level deeper
Usage by app and member
Conversations or runs, time spent, tokens and active members
Daily reports
Export cost reports by employee and by app, in USD or CNY
Model prices you maintain
Change a price and recalculate past costs
FAQ



