docs(cli): refresh choosing-your-model stack rank for June 2026#1208
Conversation
Updates the stack rank header to June 2026 and adds Claude Fable 5 (rank 1) and Claude Opus 4.8 Fast Mode (rank 2). Updates Claude Opus 4.7 (now rank 3) to drop the promotional multiplier and reflect Max and Fast reasoning at a 2x multiplier. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
|
Droid finished @factory-sarah's task —— View job The June 2026 stack-rank update looks coherent, but Section 4’s reasoning-effort guidance is now inconsistent with the new Opus 4.8 recommendation. Fixing that single mismatch should prevent readers from missing the correct reasoning settings for the top-ranked model. |
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Workflows to automatically generate PRs for you. |
| | 18 | **Droid Core (Kimi K2.6)** | Open-source, 0.4× multiplier with image support and optional High reasoning; good for cost-sensitive work when you still want a thinking toggle. | | ||
| | 19 | **Droid Core (Kimi K2.5)** | Open-source, 0.25× multiplier with image support; older Kimi option for cost-sensitive work. | | ||
| | 1 | **Claude Fable 5** | First public release Mythos model; top pick for multi-step autonomous engineering work. 4× multiplier. | | ||
| | 2 | **Claude Opus 4.8 Fast Mode** | Newest Anthropic flagship with **Max** reasoning; top pick for hardest work. | |
There was a problem hiding this comment.
[P2] “Reasoning effort settings” omits Opus 4.8 despite Section 1 claiming Max support
In the June 2026 stack-rank table, Claude Opus 4.8 Fast Mode is described as having Max reasoning, but Section 4’s model list (and the “Max is available on...” sentence) does not mention Opus 4.8. Updating Section 4 to include Opus 4.8 will keep the reasoning-effort guidance consistent with the new top recommendation.
Adds Claude Fable 5 (`claude-fable-5`, 4x multiplier, full reasoning ladder up to Max) at the top of the Anthropic section, matching the new rank 1 entry on the choosing-your-model guide. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Adds an asterisk next to Claude Fable 5 on the choosing-your-model guide and the models reference page, with a footnote linking to Anthropic's Mythos-class data retention article. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Mirrors the Opus 4.6 / 4.6 Fast pattern: each Opus generation now has a base row (no multiplier called out, 2x baseline) plus a Fast row that calls out the higher multiplier (4x for 4.8 Fast, 12x for 4.7 Fast). Renumbers ranks 2-23 accordingly. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Calls out 2x on the Opus 4.7 and Opus 4.8 base rows to sit alongside the 4x and 12x multipliers on their Fast variants. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Updates
docs/cli/user-guides/choosing-your-model.mdx:Note: the "Match the model to the job" table and the Tip block under it still reference Opus 4.7 with the old promotional 1× multiplier through April 30, and the page intro still says "last updated on Wednesday, June 3rd 2026". Those weren't part of this scope — happy to follow up in a separate PR if you'd like them refreshed too.