feat(providers): add Claude Opus 5 model

[waleedlatif1]

Summary

• Add claude-opus-5 to the Anthropic provider in models.ts — Anthropic's new Opus flagship (launched July 24, 2026)
• 1M context, 128k max output, $5/$25 per MTok, $0.50/M cached input, full effort ladder (low/medium/high/xhigh/max, default high), native structured outputs
• Marked recommended: true and moved recommended off claude-opus-4-8 (Opus 5 supersedes it as the Opus flagship)
• Every value verified against Anthropic's live docs (models overview, pricing, effort, release notes) — no temperature capability (rejected on Opus 4.7+), matching the Opus 4.8 sibling

Type of Change

• New feature (model addition)

Testing

• bun run lint clean
• providers/models.test.ts + providers/utils.test.ts pass (179 tests) in a full env

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing
• No new warnings introduced
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
docs	Ready	Preview, Comment	Jul 24, 2026 5:35pm

[cursor]

PR Summary

Low Risk
Additive model registry and Anthropic thinking classification aligned with existing Opus 4.8 behavior; no auth, billing, or core execution path refactors.

Overview
Adds claude-opus-5 to the Anthropic model catalog with pricing, 1M context, 128k max output, native structured outputs, and adaptive thinking (effort ladder through max, default high, summarized streaming). claude-opus-5 is marked recommended; recommended is removed from claude-opus-4-8 so Opus 5 is the Opus flagship in pickers.

supportsAdaptiveThinking in the Anthropic provider now treats opus-5 like Fable/Sonnet 5 and Opus 4.7/4.8 (adaptive only, no budget_tokens). Agent-events runs request display: 'summarized' for this model; thinking tests cover claude-opus-5. Agent block docs list claude-opus-5 under Anthropic “summaries only” streamed thinking.

^{Reviewed by Cursor Bugbot for commit 67946fe. Configure here.}

[greptile-apps]

Greptile Summary

Adds Claude Opus 5 support across the Anthropic provider.

• Registers model capabilities, pricing, context limits, and recommendation metadata.
• Routes Opus 5 through adaptive thinking and adds regression coverage.
• Documents summarized-thinking streaming support.

Confidence Score: 5/5

The PR appears safe to merge.

No blocking failures remain in the fix for Claude Opus 5 adaptive thinking.

Important Files Changed

Filename	Overview
apps/sim/providers/models.ts	Registers Claude Opus 5 and updates the recommended Opus flagship.
apps/sim/providers/anthropic/core.ts	Classifies Claude Opus 5 as an adaptive-thinking model.
apps/sim/providers/anthropic/core.thinking.test.ts	Extends adaptive-thinking tests to cover Claude Opus 5.
apps/docs/content/docs/en/workflows/blocks/agent.mdx	Documents Claude Opus 5 summarized-thinking streaming behavior.

_{Reviews (3): Last reviewed commit: "docs(agent): regenerate agent-stream doc..." | Re-trigger Greptile}

[waleedlatif1]

@greptile

[waleedlatif1]

@cursor review

[cursor]

✅ Bugbot reviewed your changes and found no new issues!

Comment @cursor review or bugbot run to trigger another review on this PR

^{Reviewed by Cursor Bugbot for commit ae1345e. Configure here.}

[waleedlatif1]

@greptile

[waleedlatif1]

@cursor review

[cursor]

✅ Bugbot reviewed your changes and found no new issues!

Comment @cursor review or bugbot run to trigger another review on this PR

^{Reviewed by Cursor Bugbot for commit 67946fe. Configure here.}