October 30, 2026
Heard in AI

A story we follow

Anthropic releases Claude Opus 5.5

Tracks Anthropic's September 2026 release of Claude Opus 5.5: pricing and cost against Fable 5.1 and Opus 5, capabilities such as procedurally generated media, user experience and adoption.

A story page follows one specific event across podcast discussions, with an overview and a timeline of what changed. It updates when new episodes discuss the event, and you can get those updates by email or push. How our formats work

Overview

Anthropic released Claude Opus 5.5 on September 22, 2026. It says the model performs near Fable 5.1 on most work, at $4 and $20 per million input and output tokens, and estimates typical costs about 40 percent below Opus 5. In one internal translation task, Opus 5.5 cost 51 percent less than Fable 5.1.

On Moonshots (published September 28), Peter Diamandis described this as Fable 5.1 performance at half the price, a figure that matches only that one task. A panelist called it the first truly competent model and contrasted it with an unhinged Opus 5. Dave Blundin praised its visual mock-ups, and Alex Wissner-Gross highlighted procedurally generated video and images as a code-first strategy, unlike OpenAI's pixel models. Panelists also called the release, coming days after Amodei's pacing essay, hypocritical.

On Latent Space (published October 21), a Tenex team member gave an adopter's account. He said Opus 5.5 is now the main driver for most of Tenex's engineers, often paired with Codex as reviewer, and that many people are moving away from Fable because Opus is good, fast and cheap enough to cover the range he once needed two or three models for. He credited the period from Opus 4.5 to Opus 5.5 with roughly doubling engineer output, and said Opus 5.5 had cut Tenex's token costs sharply. He argued that many economically valuable tasks no longer justify a much pricier model, though he said much more expensive, clearly better models could pull spending back up. These are one firm's observations, not measurements.

Recent coverage

Changes to our coverage published in the past seven days. These are site update dates; podcast discussion dates and sources are in each version.

No changes to this account were published in the past seven days.

What changed

Dates show when each podcast discussion was published.

  1. New information

    A Tenex engineer on Latent Space said Opus 5.5 is now the default driver for most of the team, with Codex as reviewer, and that people are moving away from Fable. He credited Opus 5.5 with sharply lower token costs and, by his own estimate, a roughly 2x rise in engineer output since Opus 4.5.

  2. New information

    Moonshots discussed the Opus 5.5 release. Diamandis said it matches Fable 5.1 at half the price and that he switched to it. A panelist called it the first truly competent model. Blundin praised its visual mock-ups. Wissner-Gross described procedurally generated video as a code-first strategy that contrasts with OpenAI's pixel models.

Podcast discussions

Sources

  1. 01
  2. 02
  3. 03

Our coverage

How Tenex keeps its coding agents' context in the repository

Tenex, which builds AI systems for large companies, says it stores plans, agent decision logs and project context beside its code so agents can read them. In the episode, Tenex said Opus 5.5 is the main coding model for most of its team, and that pairing it with Codex as a reviewer still gives the best results.

7 min read

Version history

  • Oct 30, 2026 · Version 2

    A Tenex engineer on Latent Space said Opus 5.5 is now the default driver for most of the team, with Codex as reviewer, and that people are moving away from Fable. He credited Opus 5.5 with sharply lower token costs and, by his own estimate, a roughly 2x rise in engineer output since Opus 4.5.