cloudflare / cloudflare/cloudflare-docs
Add pricing details for @cf/meta/llama-3.1-8b-instruct-fast
- Dominant language
- MDX
- Stars
- 5.2k
- Forks
- 16.7k
- Avg merge
- 2d 6h
- Merged PRs (30d)
- 337
Description
### Existing documentation URL(s)
https://developers.cloudflare.com/workers-ai/platform/pricing/
https://developers.cloudflare.com/workers-ai/models/llama-3.1-8b-instruct-fast/
### What changes are you suggesting?
The model catalog lists `@cf/meta/llama-3.1-8b-instruct-fast` as an active Cloudflare-hosted model, but the Workers AI pricing page does not list unit pricing for this exact model ID. It only lists the distinct `@cf/meta/llama-3.1-8b-instruct-fp8-fast` variant.
Please document:
1. The exact per-million input and output token prices for `@cf/meta/llama-3.1-8b-instruct-fast`.
2. Whether cached input has a separate rate.
3. Whether this model ID is internally aliased to another model for billing, and, if so, which exact pricing row applies.
4. Which returned usage fields should be used to calculate the authoritative request cost.
This clarification is needed to apply a strict pre-request cost cap in a small, controlled Workers AI benchmark. We do not want to substitute the similarly named FP8 variant's price without confirmation.
Expected outcome: add an exact pricing row for this model ID, or add an explicit note that identifies the authoritative billing alias, rates, and usage fields.
### Additional information
No account-specific data, API keys, request logs, or production payloads are involved in this documentation request.
Contributor guide
Assessment
This issue has not been assessed yet.