Conversation
Users can now select Anthropic's `claude-fable-5-1`, the successor to Claude Fable 5 for demanding reasoning and long-horizon agentic work. It carries a 1M token context window, 128k max output tokens, always-on adaptive thinking with the full effort ladder up to `max`, and a June 2026 knowledge cutoff. This repository's `fable` alias resolves to it. Like Fable 5, the model rejects both `thinking: disabled` and a forced `tool_choice` with a 400 error. Marking it always-on in the override table routes it through the existing soft-force path, so `jp query -u run_tests` downgrades to `tool_choice: auto` plus a system prompt nudge backed by escalating-nudge retries, and JP never sends a disabled thinking config the API would refuse. The model also binds each thinking block to the conversation prefix that produced it, so a request that rebuilds the system prompt between turns is rejected once Anthropic enforces the check for an account. JP's thinking-rejection recovery catches that error and replays the reasoning as `<think>` text, at the cost of one extra round trip. Opting out of the failure with `prefix_mismatch_behavior: "drop_block"`, along with per-message effort, turn-scoped system messages, and `thinking.display: "updates"`, needs request fields `async-anthropic` does not yet model. Signed-off-by: Jean Mertz <git@jeanmertz.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Users can now select Anthropic's
claude-fable-5-1, the successor to Claude Fable 5 for demanding reasoning and long-horizon agentic work. It carries a 1M token context window, 128k max output tokens, always-on adaptive thinking with the full effort ladder up tomax, and a June 2026 knowledge cutoff. This repository'sfablealias resolves to it.Like Fable 5, the model rejects both
thinking: disabledand a forcedtool_choicewith a 400 error. Marking it always-on in the override table routes it through the existing soft-force path, sojp query -u run_testsdowngrades totool_choice: autoplus a system prompt nudge backed by escalating-nudge retries, and JP never sends a disabled thinking config the API would refuse.The model also binds each thinking block to the conversation prefix that produced it, so a request that rebuilds the system prompt between turns is rejected once Anthropic enforces the check for an account. JP's thinking-rejection recovery catches that error and replays the reasoning as
<think>text, at the cost of one extra round trip. Opting out of the failure withprefix_mismatch_behavior: "drop_block", along with per-message effort, turn-scoped system messages, andthinking.display: "updates", needs request fieldsasync-anthropicdoes not yet model.