> ## Content Index > Fetch the complete content index at: https://www.testingcatalog.com/llms.txt > Use this file to discover other available public pages before exploring further. # Anthropic prepares to release Claude Fable 5.5 next week - URL: https://www.testingcatalog.com/anthropic-several-models-coming-fable-5-5-reaches-partners/ - Published: 2026-10-17T12:52:59.000Z - Updated: 2026-10-17T12:52:59.000Z - Description: Claude Fable 5.5 has reportedly reached Anthropic's partners, and six early access codenames spotted in Devin CLI, plus one in a GitHub model picker, suggest it may not be the only model on the way. - Author: Dwaybe Cubbins - Tags: Anthropic, Claude News, Latest AI News, AI Rumours Anthropic will likely serve more than one model soon, and Claude Fable 5.5 may be first in line. Recent leaks point the same way: partners appear to be testing Fable 5.5, while new Claude codenames keep turning up in third-party tools. One recent sign that Fable 5.5 may be nearing release came from X user Alex Getman, who shared a screenshot showing Fable 5.5 in Azure's East US 2 model catalog with the status set to "Generally available." The screenshot showed a LiteLLM entry for the model with the same pricing as Fable 5.1, $10 per million input tokens and $50 per million output tokens, along with a 1 million token context window and up to 128K tokens of output. > ANTHROPIC 🔥: Claude Fable 5.5 traces have been spotted in Azure configs. > > If this model drops today, it would be a new SOTA across many areas. > > Soon? 👀 [https://t.co/2T80olPKIs](https://t.co/2T80olPKIs?ref=testingcatalog.com) [pic.twitter.com/ECepatTkLz](https://t.co/ECepatTkLz?ref=testingcatalog.com) > > — 🚨 AI News | TestingCatalog (@testingcatalog) [October 13, 2026](https://x.com/testingcatalog/status/2110031026026332551?ref%5Fsrc=twsrc%5Etfw&ref=testingcatalog.com) That said, we'd take it with a grain of salt, since most of those values appear to have been carried over from Fable 5.1, and LiteLLM removed the entry a day later when Azure queries came back empty. But this isn't the only source that suggests we might see Anthropic's upcoming frontier model soon. A couple of days ago, Lyra (leaker) also said partners had been given access to the upcoming model. There's no release date yet, though the leaker expects one next week. > Partners got access to the upcoming Claude Fable 5.5. > No info on any release date this time but I assume next week. > > — lyra (@lyraxana) [October 15, 2026](https://x.com/lyraxana/status/2110732706443067733?ref%5Fsrc=twsrc%5Etfw&ref=testingcatalog.com) The same day, Lumina Bench found a "claude-onigiri-eap" codename in Artificial Analysis's frontend, with four effort levels and an Opus 4.8 fallback, a setup typical of Fable and Opus-tier models. It was pulled hours later. Lumina Bench also heard that Claude Code is routing to Fable 5.5 again, which matches what users saw last weekend, when [Opus 5.5 started sending some requests to a stronger, unannounced model](https://www.testingcatalog.com/anthropic-tests-new-model-opus-5-5-routing-opus-5-6-fable-5-5/). Back then, it could have been Fable 5.5 or an Opus 5.6 checkpoint (Highly speculative). But Fable 5.5 seems more likely. > 🚨Fable 5.5's codename might be "claude-onigiri-eap" > > I found a new codename in Artificial Analysis's frontend called "claude-onigiri-eap" > > It has 4 variants low/med/high/xhigh, it also falls back to Opus 4.8 (which Fable/Opus tier models do) > > I haven't seen this codename before… [pic.twitter.com/00P55mWZ66](https://t.co/00P55mWZ66?ref=testingcatalog.com) > > — Lumina (@LuminaBench) [October 15, 2026](https://x.com/LuminaBench/status/2110672726075535404?ref%5Fsrc=twsrc%5Etfw&ref=testingcatalog.com) As we mentioned earlier, however, Fable 5.5 might not be the only new model Anthropic will release soon. A post from [Haleemah](https://x.com/Haleeeemahh/status/2111093598649204854?ref=testingcatalog.com) showed two names spotted in Cognition's Devin CLI: claude-orzata-eap and claude-vanilla-eap. Even user AiBattle on X then mapped six early access names by build. Orzata first showed up on September 21; Boba and Vanilla on September 29; Odango on October 2; and Verrine and Onigiri on October 10. > Some new Claude EAP model names have been spotted in the Devin CLI binary > > "Verrine" and "Onigiri" seem to have been added about a week ago > > Reliabe leakers have said that partners got access to Fable 5.5 recently. A release next week seems likely [https://t.co/DuhA1GZRFN](https://t.co/DuhA1GZRFN?ref=testingcatalog.com) [pic.twitter.com/gFGLvL1kzk](https://t.co/gFGLvL1kzk?ref=testingcatalog.com) > > — AiBattle (@AiBattle\_) [October 16, 2026](https://x.com/AiBattle%5F/status/2111169488481825051?ref%5Fsrc=twsrc%5Etfw&ref=testingcatalog.com) Odango is the most interesting of the bunch because it didn't stay inside Devin. Damageboy shared a screenshot of what looks like a GitHub Copilot-style agent prompt, where "Claude Odango EAP" sat in the model picker as a selectable option, right between Claude Fable 5.1 and Claude Sonnet 5.5\. The screenshot was allegedly shared by the GitHub account on X, but the post appears to have been taken down at the time of writing. > Did Github accidentally leak a an upcoming Claude Odango (probably a codename): [https://t.co/loXPKnbGuR](https://t.co/loXPKnbGuR?ref=testingcatalog.com) [pic.twitter.com/D3x4TrFBO5](https://t.co/D3x4TrFBO5?ref=testingcatalog.com) > > — damageboy (@damageboy) [October 15, 2026](https://x.com/damageboy/status/2110590514290385226?ref%5Fsrc=twsrc%5Etfw&ref=testingcatalog.com) Of course, it's hard to match the codenames to the models since Anthropic hasn't made anything official yet. Still, Onigiri is the favorite for Fable 5.5, while the rest could be anything from Fable variants to the Opus 5.6 people keep speculating about. On the product side, recent weeks brought hints of [Claude Health for web and a mysterious "Iggy"](https://www.testingcatalog.com/anthropic-tests-claude-health-for-web-and-mysterious-iggy/), a hidden Voice tab suggesting Anthropic [might be building its own voice models](https://www.testingcatalog.com/anthropic-might-be-developing-its-own-voice-models/), and the beta launch of [Claude Dashboards and Motion](https://www.testingcatalog.com/claude-launches-dashboards-and-motion-in-beta/). This week also brought a new sessions hub in Claude Code Desktop and general availability for Code Review. Overall, Anthropic is showing no signs of slowing down. The company [launched Opus 5.5 less than a month ago](https://www.testingcatalog.com/anthropic-launches-claude-opus-5-5-with-lower-api-costs/), at $4 and $20 per million input and output tokens, roughly matching Fable 5.1 on most tasks. Then they also dropped [Sonnet 5.5](https://www.testingcatalog.com/anthropic-launches-claude-sonnet-5-5-with-faster-output/) and [Haiku 5.5](https://www.testingcatalog.com/anthropic-launches-claude-haiku-5-5-at-lower-api-prices/) shortly after. Anthropic hasn't confirmed anything. But if partners really have it, a launch next week wouldn't surprise anyone.