16 September 2026
Heard In AI
Daily email

A story we follow

Anthropic Releases Fable 5.1 and Restricted Mythos 5.1

Tracks the paired model release, access distinctions, benchmark results and enterprise inference economics.

A topic page follows one event across podcast discussions, with an overview and a timeline of what changed. It updates when new episodes discuss the event. How our formats work

Updated

Overview

Anthropic released broadly available Fable 5.1 alongside restricted Mythos 5.1, described as sharing underlying intelligence with different safety envelopes. Moonshots highlighted strong expert-reasoning results and 75 percent cheaper cache reads, arguing that reusable business context improves enterprise economics and responsiveness. Box founder Aaron Levie later said that on Box's document-centric Complex Work Eval and internal holdback tests, model quality mostly tracks coding leaderboards, with Gemini stronger than its coding reputation on some knowledge-work tasks, and that Fable 5.1 was clearly the best model Box had seen, in a tight race with Grok, Muse, the Fable class, and GPT-5.6.

What changed

Dates show when each podcast discussion was published.

  1. Aaron Levie said Box's enterprise document evals found Fable 5.1 clearly state of the art for Box's use case. He said overall model ranking still closely follows coding performance, except that Gemini is disproportionately strong on some knowledge-work tasks, and that Grok, Muse, the Fable class, and GPT-5.6 remain in a tight race.

  2. The panel presents Fable 5.1 as a leading general-purpose model and connects cheaper cached context to faster, more practical enterprise workflows.

Podcast discussions

Sources

  1. 01
  2. 02
  3. 03
  4. 04

Our coverage

AI & Money Briefing

Box's Aaron Levie expects open-weight tokens and closed-model revenue to grow together

On Training Data, Box CEO Aaron Levie describes how his customers actually pick models: a default for asking questions of their files, and hard-nosed accuracy evaluations for the high-volume extraction work where most tokens are spent. He endorses Decagon founder Jesse Zhang's argument that mature workflows migrate to open-weight models, and explains why the big labs' revenue and open-weight token volume can climb at the same time.

7 min read
AI & Work Briefing

Box's two rules for software in the agent era: beat the generic agent, then let it in

On Sequoia's Training Data podcast, Box CEO Aaron Levie said any company sitting on customers' data now has two obligations: build an agent measurably better than an off-the-shelf one at its own workflows, and expose the same capabilities to outside assistants like Claude and ChatGPT. He described the tuned search-and-retrieval harness behind Box's agent, the evaluations that track model progress, and his bet that within five years roughly 90% of enterprise tokens will be spent on work nobody asked for directly.

8 min read
AI & Money Briefing

Anthropic's cheaper cached reads make business context the prize

Anthropic's Fable 5.1 charges $0.25 per million tokens for cached reads, a quarter of the previous rate, which one Moonshots panelist read as an invitation to load an entire company's context into the model and keep it there. The panel connected that price to a wider scramble: with model leads lasting about a month, the labs are racing to convert them into customer workflows, partnerships and proprietary design data that a rival cannot copy.

5 min read

Version history

  • 16 Sep 2026 · Version 2

    Aaron Levie said Box's enterprise document evals found Fable 5.1 clearly state of the art for Box's use case. He said overall model ranking still closely follows coding performance, except that Gemini is disproportionately strong on some knowledge-work tasks, and that Grok, Muse, the Fable class, and GPT-5.6 remain in a tight race.

Your privacy

Necessary

Keep forms secure, remember your cookie choice and support editor sign-in.

Always on

Help us see which articles get read and whether readers return. With your permission, we use a random browser ID. We store statistics on our own server for up to 180 days.

Refusing analytics will not change your access to articles or the daily email.

Cookie details · Privacy