Plans and Concurrent Unit Limits

Unlike most serverless offerings that bill by tokens (and create in monthly billed amounts) Featherless plans are effectively reservations of capacity, determining your model size limit and concurrent unit budget.

Individual plans include a fixed concurrent unit budget for interactive use. Business plans make that budget explicit and scalable, from a few concurrent units to as many as your workload needs.

Model Concurrent Unit Costs

Each model has a concurrent unit cost based on its size.

When a request starts, it reserves that model's concurrent unit cost from your plan budget. The units are released when the request completes or is cancelled, so higher-cost models reduce the number of requests that can run in parallel.

Plans and Allotments

Feather Chat

• Concurrent Units Included: 4

• Model size limit: no limit (i.e. DeepSeek, Kimi & GLM included)

• What You Can Run:

  • 4 concurrent units to models ≤15B, OR

  • 2 concurrent units to models ≤34B, OR

  • 1 concurrent unit to any model ≥70B, OR

  • Any combination with total concurrent unit cost up to 4

Feather Scale

• Concurrent Units Included: 8 per scale unit

• Model Size Limit: 72B (larger models excluded)

• What You Can Run:

  • 8 concurrent units to models ≤15B, OR

  • 4 concurrent units to models ≤34B, OR

  • 2 concurrent units to models ≤70B, OR

  • Any combination with total concurrent unit cost up to 8

Examples

Feather Chat Examples

1. Example 1: Run 4 simultaneous requests to Qwen2.5-7B (≤15B)

2. Example 2: Run 2 simultaneous requests to Qwen2.5-32B (24-32B)

3. Example 3: Run 1 simultaneous request to Qwen2.5-72B (≥70B)

4. Example 4 (Mixed): Run 1 Qwen2.5-32B (cost 2) + 2 Qwen2.5-7B (cost 1 each) = total concurrent unit cost 4

Feather Scale Examples

1. Example 1: Run 8 simultaneous requests to Qwen2.5-7B (≤15B)

2. Example 2: Run 4 simultaneous requests to Qwen2.5-32B (24-32B)

3. Example 3: Run 2 simultaneous requests to Qwen2.5-72B (≥70B)

4. Example 4 (Mixed): Run 1 Qwen2.5-72B (cost 4) + 2 Qwen2.5-32B (cost 2 each) = total concurrent unit cost 8

Combinations

Your total concurrent unit usage is calculated by adding the concurrent unit cost of all models running at the same time. For example:

• 1 model with cost 2 + 2 models with cost 1 = total concurrent unit cost 4

• 1 model with cost 4 + 2 models with cost 2 = total concurrent unit cost 8

Requests that would take the total concurrent unit cost above the subscriber’s purchased allotment are rejected with HTTP status code 429.