# llm-anthropic 0.31 adds Fable 5.2 and Mythos 5.2; max effort returns no image

> The llm-anthropic 0.31 plugin adds Anthropic's Fable 5.2 and Mythos 5.2, but a max reasoning effort test ran out of tokens without returning an image.

- Status: reported (One outlet, not independently verified yet.)
- Topic: Products
- Published: 2026-11-09T23:58:46+00:00
- Updated: 2026-11-10T01:08:59.819488+00:00
- Web page: https://ai-news.ikangai.com/story/llm-anthropic-0-31-adds-fable-5-2-and-mythos-5-2-max-effort-returns-no-image/
- Seen in: Simon Willison
- Change: capability (New capability)
- Applies to: llm-anthropic 0.31, Fable 5.2, Mythos 5.2, claude-fable-5.2

## What changed

The llm-anthropic 0.31 plugin adds support for Anthropic's Fable 5.2 and Mythos 5.2 models, and max reasoning effort on Fable 5.2 ran out of tokens without returning an image in Willison's test.

Simon Willison's llm-anthropic 0.31 plugin adds support for Anthropic's Fable 5.2 and Mythos 5.2 models. His test of the new Fable model turned up a failure at the highest reasoning setting. Willison ran his SVG pelican prompt at five reasoning levels: low, medium, high, xhigh and max. The max run never returned an image. It spent 128,000 tokens thinking and ran out of tokens before producing output. He describes this as the same bug he saw with Opus 5.5, but new to the Fable series, since Fable 5 and Fable 5.1 both returned max-effort pelicans without trouble.

Running the model is a short command-line task. Install the llm tool and the llm-anthropic plugin, store an Anthropic key, then pass claude-fable-5.2 as the model name with a prompt. That is enough to generate an SVG. Willison's post links to all five pelicans, and he picked the xhigh result as the best of them. The failure matters because a token budget that runs out during thinking leaves the user with no image at all.

Anthropic's announcement, published the same day, describes Fable 5.2 and Mythos 5.2 as upgrades for the most demanding technical work. The company says the two models are the same underlying model. Fable has stronger cybersecurity and biology safeguards so it can be offered to all users. Organizations with beneficial cybersecurity or biomedical research aims can apply for access to Mythos through its Life Sciences and Cyber Verification Programs. Anthropic lists pricing at $8 per million input tokens and $40 per million output tokens. It says outputs come up to 30% faster than Fable 5.1. An optional fast mode costs $16 and $80 per million tokens and speeds generation by up to 2.5 times.

The announcement's benchmark table compares Fable 5.2 with Fable 5.1, Opus 5.5, GPT-6 Astra and Gemini 4 Carbon. Its scores are tied to the effort setting each model ran at. Anthropic's own FrontierCode 2.0 figures show Fable 5.2 at 45.3% on High effort, roughly level with Opus 5.5's best 45.5%, and 49.1% at Xhigh. The post does not say how the max setting performed on that benchmark. It also does not address whether max effort can fail to return output in the way Willison saw.

## Why it matters

Teams that set maximum reasoning effort on Fable 5.2 may get empty outputs after heavy token use, which can break pipelines that expect a response.

## What you can do

Test the max reasoning setting on Fable 5.2 with your own prompts, and set token limits before using it in production.

## Original source

- [Release 0.31 · simonw/llm-anthropic · GitHub](https://github.com/simonw/llm-anthropic/releases/tag/0.31)
- [Release: llm-anthropic 0.31](https://simonwillison.net/2026/Nov/9/llm-anthropic/)
- [Introducing Claude Fable 5.2 and Claude Mythos 5.2 \ Anthropic](https://www.anthropic.com/claude-fable-and-mythos-5-2)

---

AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it.

Cite as: IKANGAI AI News, "llm-anthropic 0.31 adds Fable 5.2 and Mythos 5.2; max effort returns no image", 2026-11-09, https://ai-news.ikangai.com/story/llm-anthropic-0-31-adds-fable-5-2-and-mythos-5-2-max-effort-returns-no-image/
