@mlapierre Thanks for this issue!

1. Update documentation to reference the Cloud Run deployment scripts as the recommended approach for load testing

Can Cloud Run deployment scripts be used for load testing already? If we remove the Runway deployment now, we would only use the ability to run load tests in the CI, correct?

FYI @sarah_zywicki

What does this MR do and why?

Updates Sonnet 4.6 and Opus 4.6 details. Both models will be charged at the standard rate:

From Anthropic

The 1M token context window is now generally available for Claude Opus 4.6 and Sonnet 4.6 via API. Both models include the full window at standard pricing—$5/$25 per million tokens for Opus 4.6 and $3/$15 for Sonnet 4.6. Previously there were separate rate limits above and below 200K tokens. We’ve simplified this to a single rate limit for the full context window. As part of this, we’ve raised your base rate limit on Opus 4.6 to 18M to accommodate your existing long context usage with room to grow. Note this applies only to the Gitlab Production (managed account) org. As always, please reach out to request additional increases.

References

• gitlab-org/modelops/applied-ml/code-suggestions/ai-assist#2038

Screenshots or screen recordings

N/A

How to set up and validate locally

MR acceptance checklist

Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.

Martin Wortschack (0b190f49) at 17 Mar 06:36

It was already added to the merge train but the pipeline failed

@igor.drozdov Can you please do the maintainer review?

@junminghuang I've addressed your comments / replied. Mind taking another look?

I agree but I'm not sure if the model configuration is the best place for this information. The Anthropic announcement is already linked in the MR description 🤔

Yeah makes sense, I've updated for other occurrences in this file as well for consistency.

Martin Wortschack (d1295b99) at 16 Mar 16:24

@junminghuang

1. Are we already able to filter llm request when doing usage billing (bill customer / not bill)? If yes, how can we usage such feature? If not, do you know if it is something already on the usage billing roadmap and how complex is that?

I don't think there is any filter logic on AIGW/DWS - we emit usage billing events for all features. There is a white-list approach on CustomersDot (CDot) to decide what features are billable. Usage billing events follow the format in https://gitlab.com/gitlab-org/modelops/applied-ml/code-suggestions/ai-assist/-/blob/main/docs/billing_events.md#trigger-events

Imho compaction LLM requests should be identified and tagged in AIGW/DWS. Filtering them out of billing events should happen upstream in CDot, this is also consistent with how other events are excluded from usage billing.

@junminghuang Mind reviewing this change?

Martin Wortschack (9f46182f) at 16 Mar 16:08

What does this merge request do and why?

Now that long-context window is GA for Sonnet 4.6 and Opus 4.6 (see https://claude.com/blog/1m-context-ga), context-1m-2025-08-07 beta headers are no longer needed for Sonnet 4.6 and Opus 4.6

From Anthropic:

The 1M token context window is now generally available for Claude Opus 4.6 and Sonnet 4.6 via API. Both models include the full window at standard pricing—$5/$25 per million tokens for Opus 4.6 and $3/$15 for Sonnet 4.6. Previously there were separate rate limits above and below 200K tokens. We’ve simplified this to a single rate limit for the full context window. As part of this, we’ve raised your base rate limit on Opus 4.6 to 18M to accommodate your existing long context usage with room to grow. Note this applies only to the Gitlab Production (managed account) org. As always, please reach out to request additional increases.

We are also updating pricing multipliers in https://gitlab.com/gitlab-org/customers-gitlab-com/-/merge_requests/15070

Numbered steps to set up and validate the change are strongly suggested.

Merge request checklist

• Tests added for new functionality. If not, please raise an issue to follow up.
• Documentation added/updated, if needed.
• If this change requires executor implementation: verified that issues/MRs exist for both Go executor and Node executor or confirmed that changes are backward-compatible and don't break existing executor functionality.

Martin Wortschack (2697b790) at 16 Mar 07:03

Martin Wortschack (8b78b9e4) at 16 Mar 06:58

Martin Wortschack (cc0a03c6) at 16 Mar 06:58

... and 1 more commit

Why is this change being made?

Updates links on https://handbook.gitlab.com/handbook/engineering/ai/ai-framework/#-backlog-refinement

CHANGEME

Author and Reviewer Checklist

Please verify the check list and ensure to tick them off before the MR is merged.

• Provided a concise title for this Merge Request (MR)
• Added a description to this MR explaining the reasons for the proposed change, per say why, not just what
  • Copy/paste the Slack conversation to document it for later, or upload screenshots. Verify that no confidential data is added, and the content is SAFE
• Assign reviewers for this MR to the correct
  • The when to get approval handbook section explains when DRI approval is required
  • The who can approve handbook section explains how to identify the DRI
  • If the MR does not require DRI approval, consider asking someone on your team, such as your manager.
  • The approver may merge the MR. If they approve but don't merge, you can merge.
• For transparency, share this MR with the audience that will be impacted.
  • Team: For changes that affect your direct team, share in your group Slack channel
  • Department: If the update affects your department, share the MR in your department Slack channel
  • Division: If the update affects your division, share the MR in your division Slack channel
  • Company: If the update affects all (or the majority of) GitLab team members, post an update in #whats-happening-at-gitlab linking to this MR
    • For high-priority company-wide announcements work with the internal communications team to post the update in #company-fyi and align on a plan to circulate in additional channels like the "While You Were Iterating" Newsletter

Commits

• AIF handbook: Update issue refinement links

---