Two unlisted Claude model identifiers surfaced in developer tools and on Discord in late August 2026, and the speculation they triggered did not stay speculation for long. On September 1, 2026, Anthropic released Claude Fable 5.1 and Claude Mythos 5.1, confirming that a double release really was coming even as questions remain about how those two shipped models connect to the leaked identifiers. Add in a separate report that Anthropic was quietly benchmarking a frontier system well past its flagship at the time, and you get the full arc of a story that started as a rumor and closed out with an actual ship date within days.
Before the September 1 release, neither leaked identifier had a public system card, a pricing page, or a confirmed ship date, and the pattern was already hard to ignore: developers had spotted new API identifiers in the wild, Anthropic’s own August risk report quietly named an internal system that beat its newest public model at the time, and industry chatter about a possible $100 billion Anthropic valuation had put every scrap of roadmap information under a microscope. Here is what actually shipped, what is still unconfirmed about how the leaked identifiers map onto it, and why a genuinely simultaneous double release was such a departure from how Anthropic shipped models for most of the year.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
The Leak: Two New Model IDs Show Up in the Wild
The story starts with a pair of strings that Anthropic never announced. According to Times of AI, two new identifiers, claude-marshmallow-eap and claude-melon-eap, appeared in third-party developer apps and on Discord around August 24, 2026. Neither name traces back to an official Anthropic announcement. There is no model card, no listed release date, and no entry on the public API pricing page for either one.
The “eap” suffix is the detail developers latched onto first. It typically signals an Early Access Program build, the same kind of pre-release tagging that has preceded past Claude launches. That is not proof of an imminent release on its own, but it is the strongest signal yet that Anthropic is running two separate model tracks in parallel rather than one.
claude-marshmallow-eap
claude-melon-eapPer Times of AI’s reporting, the most common theory among developers who spotted the strings is that Marshmallow corresponds to a Claude Opus 5.1 update and Melon maps to a Claude Sonnet 5.1 refresh. A smaller group of developers floated a different read: that Melon could represent an entirely new Haiku-class model rather than a Sonnet point release. Anthropic has not confirmed either theory, and the company has a track record of shipping codenames that do not map cleanly onto the tier they end up shipping in — the two models Anthropic actually released on September 1, Claude Fable 5.1 and Claude Mythos 5.1, matched neither the Opus/Sonnet nor the Haiku theory directly.
Model 2: The Internal System Anthropic Didn’t Mean to Reveal
The second thread predates the Marshmallow and Melon leak by about ten days and comes from a very different source: Anthropic itself. On August 14, 2026, the company published a 186-page company-wide risk report, and buried inside it was the first public acknowledgment of a system referred to only as “Model 2,” reported BigGo Finance. The report described Model 2 as more capable than Claude Mythos 5, Anthropic’s strongest publicly released model at the time, and stated explicitly that the company has no current plans to release it externally.
GizChina’s coverage of the same disclosure notes that no release date has been set and frames Model 2 as a research and acceleration tool being used internally rather than a product in the pipeline for near-term launch. Anthropic’s own system cards page does not list Model 2 at all, which tracks with the company’s insistence that the system is not on a public roadmap.
Separate reporting adds a number to the story. A Seeking Alpha piece on Anthropic’s reported ambitions for a $100 billion IPO said the company’s researchers are actively benchmarking Model 2, describing it as widely speculated to become a future Claude Mythos-line release. Per that report, early internal evaluations show Model 2 scoring 62.8% on Anthropic’s CoBench v2 development benchmark, compared with 50.3% for Claude Mythos 5. That is a meaningful jump on an internal yardstick Anthropic uses to measure AI research and development capability, though it is worth stressing this is an unreleased, unaudited internal number, not a published benchmark result.
Why the “Double Release” Framing Is Catching On
Individually, either story would be a minor footnote. A pair of unlisted API strings is not unusual for a company that ships as often as Anthropic does, and a research lab quietly testing a stronger internal model is standard practice at every major AI lab. What has pushed this into breaking-news territory is the overlap: two unexplained model identifiers surfacing within days of a disclosure that Anthropic is running an internal system that beats its current flagship, all inside the same two-week window in August 2026.
That overlap is why outlets including thewincentral.com and Seeking Alpha have been treating the “two new models” angle as the story worth chasing this week, rather than either leak in isolation. None of the reporting confirms that Marshmallow, Melon, and Model 2 are the same thing, or that all of it lands on the same release date. It is entirely possible Marshmallow and Melon are minor point releases while Model 2 stays internal for months longer. But the clustering of signals is exactly the kind of pattern that has preceded real Anthropic launches before, most recently the “Honeycomb” leak cycle that predicted Claude Sonnet 5, Fable 5, and Opus 5 weeks ahead of their actual ship dates.
The Honeycomb Precedent: Anthropic’s Last Leak Cycle Played Out Exactly Like This
This would not be the first time stray identifiers turned into a real Claude release inside a few weeks. According to a running roadmap tracker maintained by AI Tools Review, a set of leaked codenames dubbed the “Honeycomb” cycle correctly predicted three separate launches earlier this summer. Claude Sonnet 5 shipped on June 30 at an introductory $2 per million input tokens and $10 per million output tokens, Fable 5 and Mythos 5 were restored to availability on July 1, and Claude Opus 5 launched as the new general-purpose flagship on July 24 at $5 per million input tokens and $25 per million output tokens. All three matched names that had circulated as unconfirmed leaks in the weeks prior. Anthropic’s own system cards go further, framing Sonnet 5 and Opus 5 as new fifth-generation Claude models, and for a stretch after its June 2026 debut, Fable 5 was described as Anthropic’s most capable public model, edging out Opus 4.8 on capability until Opus 5 took over the flagship spot in late July.
That same tracker noted that, as of early August, one loose thread remained: a possible Claude Fable 5.1 update that had not yet shipped and had no system card, pricing page, or Anthropic announcement to back it up at the time. The Marshmallow and Melon identifiers that surfaced August 24 could be an extension of that same unresolved thread, or a fresh cycle entirely. Anthropic does not comment on unreleased products, so outside observers were left triangulating from API strings, developer forum posts, and whatever the company let slip in unrelated documents like its risk reports — at least until the September 1 release settled part of the question, covered next.
The Double Release Landed on September 1 — Just Not the Way Developers Guessed
The wait was short. On September 1, 2026, Anthropic released Claude Fable 5.1 and Claude Mythos 5.1, confirming that the “double release” framing circulating since late August was directionally right. Two models did ship together, just eight days after the claude-marshmallow-eap and claude-melon-eap identifiers surfaced in developer tools and on Discord around August 24. Anthropic’s own September 2026 system card labels both releases version 5.1, but access is not identical: Mythos 5.1 is limited to vetted programs, while Fable 5.1 has the broader public rollout.
What caught most of the developer speculation off guard was which lines actually got the update. The dominant theory around the leaked marshmallow and melon strings pointed to an Opus 5.1 and Sonnet 5.1 pairing, with a smaller camp betting on a new Haiku-class model. Instead, Anthropic delivered on the other loose thread this site flagged in early August: a rumored Fable 5.1 update that had no system card or ship date at the time. Anthropic paired it with a Mythos 5.1 release, extending the same two-model lineup that launched together on July 1, 2026. The company has not stated whether claude-marshmallow-eap and claude-melon-eap were internal build names for Fable 5.1 and Mythos 5.1 specifically, or whether they refer to separate systems that still have not shipped. That link remains unconfirmed even after the release.
The timing matters more than the naming did. An eight-day gap between a leaked identifier and a shipped product is faster than the turnaround during the Honeycomb cycle earlier this year, when weeks separated the first leaked codenames from Claude Sonnet 5, Fable 5, Mythos 5, and Opus 5 actually shipping. That compressed timeline suggests Anthropic’s release pipeline for the Fable and Mythos updates was already well along by the time outside observers spotted anything, and that the “eap” tag developers noticed on the leaked strings really was signaling an imminent release rather than an early, unfinished build.
For Model 2, nothing has changed. Anthropic’s system cards page still does not list it, and the September 1 announcement did not reference it by name. The internal system that reportedly scored 62.8% on CoBench v2, against 50.3% for the publicly released Claude Mythos 5, remains exactly where Anthropic said it would stay in the August risk report: internal, unreleased, and without a stated ship date. Whether Fable 5.1 and Mythos 5.1 carry over any of the capability gains referenced in that report is not something Anthropic has addressed publicly.
Claude’s 2026 Release Cadence So Far
To understand how unusual a double release once looked, it helps to see how tightly spaced Anthropic’s confirmed 2026 launches have already been. The pace was set early: Claude Opus 4.6 and Claude Sonnet 4.6 shipped as a pair of “hybrid reasoning” models in February 2026, with Opus 4.6 launching February 5 carrying a 1,000,000-token beta context window, a 90.2% score on BigLaw Bench, and coding skills Anthropic said improved meaningfully over its predecessor, and Sonnet 4.6 following on February 17 as Anthropic’s new default, closing much of the capability gap with Opus 4.6 within the same month. That default-model handoff repeated in June 2026, when Claude Sonnet 5 launched and immediately took over as the default model for Free and Pro users. The company kept shipping a new model roughly every six to eight weeks for most of the year, alternating between the Opus, Sonnet, and Mythos lines, with the gap between Opus 5 and the September 1 Fable 5.1/Mythos 5.1 release compressing to roughly five and a half weeks before Anthropic moved again within weeks, launching Claude Opus 5.5 in September 2026 as the first model in a new 5.5 family, with Anthropic saying Sonnet and Haiku versions of that same generation would follow.
| Model | Confirmed Release Date | Tier / Line | Status |
|---|---|---|---|
| Claude Opus 4.6 | February 4-5, 2026 | Opus | Superseded |
| Claude Sonnet 4.6 | February 17, 2026 | Sonnet | Superseded |
| Claude Opus 4.8 | May 2026 | Opus | Active |
| Claude Sonnet 5 | June 30, 2026 | Sonnet | Active, default Sonnet |
| Claude Fable 5 / Mythos 5 | July 1, 2026 | Mythos line | Superseded by 5.1 |
| Claude Opus 5 | July 24, 2026 | Opus | Active, current flagship |
| Claude Marshmallow / Melon (eap) | Unconfirmed | Unknown (Opus/Sonnet suspected) | Leaked; not matched to a shipped model |
| Claude Fable 5.1 / Mythos 5.1 | September 1, 2026 | Mythos line | Active, current |
| Model 2 | No release planned | Unknown, Mythos-line suspected | Internal only |
Notice the gap between Opus 5 shipping on July 24 and the Marshmallow and Melon identifiers appearing exactly a month later on August 24. That gap also sits in the middle of a broader lifecycle shift: Anthropic’s April 14, 2026 deprecation notice scheduled Sonnet 4 and Opus 4 (the model IDs ending in 20250514) for retirement in favor of Sonnet 4.6 and Opus 4.8 effective June 15, 2026, quietly closing out the 4.0-era models while the Fable and Mythos leaks were still months away. Anthropic’s actual follow-through, releasing Fable 5.1 and Mythos 5.1 together on September 1, did mark the first genuinely simultaneous dual launch of the year, even though it extended the Fable/Mythos pairing rather than the Opus and Sonnet lines most developers expected from the leaked names. Readers who want the full history of how this cadence evolved can check our earlier coverage of why Anthropic’s August already looked different from every prior month before this latest leak even surfaced.
What the Risk Report Actually Says About Model Safety
The same August risk report that named Model 2 also raised Anthropic’s own internal misalignment rating from “very low” to “low,” a change the company attributed to the increasing capability of its models rather than any specific incident. The redacted version of the report, published on Anthropic’s own site, includes model-by-model figures for a metric the company tracks internally, showing rates of 1.4% for Claude Opus 4.7, 0.27% for Claude Opus 4.8, and 2.7% for Claude Fable 5. Anthropic has not published an equivalent figure for Model 2, and the company’s Transparency Hub does not yet reference it by name.
That distinction matters for the double-release story. A rising internal risk rating, paired with a stronger unreleased model, is consistent with a company that is being more cautious about what it ships next rather than rushing two models out the door. It is possible that caution is exactly why Marshmallow and Melon, if they are indeed close to release, have not been formally announced even after their identifiers leaked.
That caution extended to Anthropic’s external testing pipeline too. Reuters reported on August 31, 2026, that Anthropic had resumed external cybersecurity testing after pausing it following July incidents involving Claude Opus 4.7, Claude Mythos 5, and an internal research test model. Anthropic did not specify whether that internal test model was Model 2 or a separate system, so the two disclosures should be treated as related in timing rather than confirmed as the same story. Resuming outside security testing one day before shipping Fable 5.1 and Mythos 5.1 suggests Anthropic wanted its external review pipeline back online ahead of a major release rather than after one.
Benchmark Data Points Fueling the Speculation
The numbers driving most of this week’s coverage come from CoBench v2, Anthropic’s internal benchmark for measuring AI research and development capability. It is not a public leaderboard metric like MMLU or SWE-bench, which makes the reported scores harder to independently verify, but it is the same benchmark Anthropic itself cited in the risk report that first named Model 2.
| System | CoBench v2 Score | Source | Availability |
|---|---|---|---|
| Claude Mythos Preview | 54.8% | Anthropic risk report (via AIBase) | Limited |
| Claude Mythos 5 | 50.3% | Anthropic risk report / Seeking Alpha | Public |
| Model 2 (internal) | 62.8% | Seeking Alpha | Not released |
| Claude Marshmallow / Melon (eap) | Not disclosed | Times of AI | Leaked identifiers only |
| Claude Fable 5.1 / Mythos 5.1 | Not yet disclosed | — | Public (shipped Sept 1, 2026) |
Two things stand out. First, Mythos Preview scoring higher than the publicly shipped Mythos 5 on this particular benchmark is a reminder that internal test builds do not always translate cleanly into shipped performance, which is a useful caveat for anyone assuming Model 2’s 62.8% figure will hold once, or if, it becomes a real product. Second, there is no published benchmark data at all for Marshmallow or Melon, since neither has a model card. Any performance claims circulating about those two identifiers right now are speculation, not measured results. Third, Anthropic has not published CoBench v2 figures for the newly shipped Fable 5.1 or Mythos 5.1 either, so there is currently no independent way to confirm whether those two updates narrow the gap with Model 2’s internal score.
The $100 Billion Backdrop
None of this is happening in a vacuum. Reports about Anthropic exploring a roughly $100 billion valuation through a potential IPO have circulated alongside the Model 2 disclosure, and it is not hard to see why the timing matters. A frontier lab preparing to raise capital or go public has every incentive to demonstrate that its research pipeline extends well past whatever it currently sells. Model 2 scoring meaningfully higher than the public flagship on an internal benchmark serves that narrative, whether or not the company intends to ship it soon.
That context also explains some of the skepticism worth applying here. A “we have something better internally” disclosure is a common move for AI labs navigating funding conversations, and it does not by itself guarantee two consumer-facing models are about to ship. Readers who want more background on how Anthropic has been positioning its market opportunity this year can see our coverage of Anthropic’s $30 trillion market claims tied to AI-run lab robots, which covers a separate but related pattern of the company talking up long-term addressable market ahead of major funding milestones.
Competitive Pressure Is Part of the Story Too
Anthropic is not shipping models in isolation. Claude Opus 4.8, which Anthropic launched on May 28, 2026 at $5 per million input tokens and $25 per million output tokens — a rate Anthropic has left unchanged since launch — with stronger coding and agentic performance and held onto as its latest Opus flagship all the way through July, still had DeepSeek’s V4-Flash-Vision-Exp release reported to come close to matching it on multimodal agent benchmarks. Every major lab, OpenAI, Google, and now several well-funded Chinese labs, is on a release cadence measured in weeks rather than quarters. The September 1 double release fits that broader industry pattern of labs shipping multiple models close together rather than one at a time, even though Anthropic paired two updates within its existing Mythos line rather than spanning separate tiers in the same announcement.
That pattern is visible in how Anthropic’s existing lineup is already priced and positioned across tiers, something we broke down in detail in our comparison of Claude Opus 5 against Mythos 5 and Gemini 3.1 Pro. The actual September 1 release did not follow the Opus 5.1/Sonnet 5.1 script most early speculation expected for Marshmallow and Melon; instead, Anthropic extended the Fable and Mythos line with 5.1 updates, its second joint release of that pairing this year.
What Developers and Enterprises Should Actually Do Right Now
That advice has partly changed. Claude Fable 5.1 and Claude Mythos 5.1 are no longer speculative; they are shipped, named models as of September 1, 2026, which puts them in a different category from claude-marshmallow-eap and claude-melon-eap. Those two EAP identifiers remain unconfirmed and still lack a documented API endpoint, a pricing tier, or a migration path, so building against them directly is still a good way to end up debugging a 404 the moment Anthropic rotates or deprecates a leaked string. Pricing itself has already moved more than once this cycle: Anthropic had floated raising Sonnet 5 to $3/$15 per million input/output tokens, then cancelled that increase in August 2026 and made the original $2/$10 pricing permanent as of August 10, 2026 — only to supersede Sonnet 5 altogether later that same month, when Anthropic launched Claude Sonnet 5.5 in September 2026, running 30% faster and costing up to 30% less than its predecessor. That is a useful reminder that even confirmed pricing, and even confirmed models, can shift before they settle. Teams evaluating the new Fable 5.1, Mythos 5.1, and Sonnet 5.5 releases should check Anthropic’s own system cards and pricing pages for current specifications rather than relying on pre-launch coverage, since pricing and benchmark details can shift between an announcement and general availability.
Teams evaluating which current model to standardize on can get a clearer picture from existing, shipped comparisons rather than speculation. Both Sonnet 5 and Opus 5 remain generally available through Anthropic’s API, priced at $2 and $5 per million input tokens respectively — Opus 5’s full $5/$25 per million input/output token rate was set at its July 2026 launch to match the pricing Anthropic had already established for Opus 4.8, rather than introducing a new tier — giving enterprises a stable baseline while pricing for the newer Fable 5.1, Mythos 5.1, and Sonnet 5.5 releases settles. Our breakdown of Claude Fable 5 against Opus 5 and GPT-5.6 Sol covers the pricing and context-window tradeoffs across last quarter’s confirmed lineup, and remains a useful reference point until Anthropic publishes fuller pricing and benchmark specifications for the new Fable 5.1 and Mythos 5.1 releases.
What’s Confirmed vs. What’s Still Rumor
It is worth separating the pieces of this story cleanly, because they carry very different levels of certainty.
- Confirmed: Two unlisted identifiers, claude-marshmallow-eap and claude-melon-eap, appeared in developer tools and on Discord around August 24, 2026, per Times of AI.
- Confirmed: Anthropic’s August 14 risk report disclosed an internal system called Model 2 that outperforms Claude Mythos 5 on the company’s CoBench v2 benchmark, per BigGo Finance and Anthropic’s own published report.
- Confirmed: Anthropic has stated it has no current plans to release Model 2 externally.
- Confirmed: Anthropic released Claude Fable 5.1 and Claude Mythos 5.1 on September 1, 2026.
- Confirmed: Anthropic resumed external cybersecurity testing on August 31, 2026, after July incidents involving Claude Opus 4.7, Claude Mythos 5, and an internal research test model, per Reuters.
- Unconfirmed: Whether claude-marshmallow-eap and claude-melon-eap were early build names for Claude Fable 5.1 and Claude Mythos 5.1, or refer to separate, still-unreleased systems.
- Unconfirmed: Whether Model 2 is the same “internal research test model” referenced in the July cybersecurity incidents.
- Unconfirmed: Any release date for Model 2.
Historical Context: How Often Leaks Turn Into Real Launches
Anthropic’s track record with leaked identifiers is mixed but leans toward eventual confirmation. Earlier in 2026, a data leak first revealed the existence of Claude Mythos months before its formal release, an episode covered at the time by Seeking Alpha. The Honeycomb cycle discussed above turned three separate leaked names into three shipped products within about a month. Not every leaked string ships, and Anthropic has quietly dropped or renamed internal projects before without ever making them public. But the base rate for this specific company, once an identifier reaches developer tooling rather than staying confined to internal Slack channels, has trended toward eventual release rather than permanent cancellation. The marshmallow and melon identifiers add another data point to that trend: within eight days of reaching developer tooling, Anthropic had a real release out the door, even if the specific product names did not match early guesses.
Market Impact: Why This Matters Beyond Claude Users
The double release landed at a moment when Anthropic’s competitive position was already under scrutiny from multiple directions: reported IPO plans that would put a hard number on the company’s valuation, rising benchmark scores from DeepSeek and other rivals, and a self-reported increase in its own model risk rating. Enterprises evaluating multi-model strategies, including those comparing Claude against Perplexity for research and retrieval workloads, now have a concrete data point rather than a hypothetical: Anthropic shipped Fable 5.1 and Mythos 5.1 simultaneously on September 1 rather than staggering them, even though the pairing came from the Mythos line rather than an Opus and Sonnet refresh. That simultaneous ship is itself a signal that Anthropic has enough compute and engineering headroom to run parallel release pipelines, which matters as much for procurement planning as it does for raw model capability.
Predictions: What Happens Next
Based on the confirmed pattern of leaks preceding launches this year, here is what is likely to happen over the coming weeks.
- That played out fast: Anthropic released Claude Fable 5.1 and Claude Mythos 5.1 on September 1, 2026, just eight days after the marshmallow and melon identifiers leaked, matching the same leak-precedes-launch pattern seen during the Honeycomb cycle. Anthropic still has not confirmed either leaked identifier by name.
- A simultaneous, same-day release did happen on September 1, which breaks from the staggered pattern Anthropic favored for most of 2026. It just was not an Opus 5.1 and Sonnet 5.1 pairing as this prediction assumed; Anthropic doubled up on the Fable and Mythos line instead, the same pairing it used on July 1.
- This has held true so far. Model 2 was not part of the September 1 announcement, consistent with Anthropic’s stated position that it has no current plans to release the system externally.
- Expect further leak-driven coverage about Model 2 specifically, since that system has not been officially confirmed for release, even after developer communities correctly anticipated the September 1 Fable 5.1 and Mythos 5.1 launch by monitoring API identifiers.
- Any formal Anthropic announcement will likely arrive alongside updated system cards and pricing, consistent with how Opus 5 and Sonnet 5 were both announced with same-day availability details across the Claude API, AWS Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.
The Bottom Line
Two things are true at once here. The double release did happen: Anthropic shipped Claude Fable 5.1 and Claude Mythos 5.1 on September 1, 2026, just eight days after the marshmallow and melon identifiers leaked, confirming that something real was moving through Anthropic’s pipeline in late August. But “two models shipped” and “the leaked identifiers were Fable 5.1 and Mythos 5.1” are not the same claim, and Anthropic has not confirmed the second one. Model 2 also remains exactly where the company said it would stay: an internal system with no system card, no pricing page, and no stated release plan. The responsible read now is that the leak-to-launch pattern held again, consistent with the Honeycomb cycle earlier this year, even though the specific mapping between the leaked codenames and the shipped products is still unconfirmed.
Frequently Asked Questions
What are claude-marshmallow-eap and claude-melon-eap?
They are two unofficial API identifiers that surfaced in third-party developer apps and on Discord around August 24, 2026, according to Times of AI. Neither has a public model card, pricing listing, or confirmed release date from Anthropic. As of the September 1, 2026 release of Claude Fable 5.1 and Claude Mythos 5.1, Anthropic still has not confirmed whether either identifier corresponds to one of those two shipped models.
Is Anthropic really preparing two new Claude models?
Yes, at least in part. Anthropic released Claude Fable 5.1 and Claude Mythos 5.1 on September 1, 2026, confirming that a double release did happen just eight days after the marshmallow and melon identifiers leaked. What is still unconfirmed is whether those two shipped models are the same systems behind claude-marshmallow-eap and claude-melon-eap, since Anthropic has not connected the leaked identifiers to the September 1 release by name.
What are Claude Fable 5.1 and Claude Mythos 5.1?
They are the two models Anthropic released on September 1, 2026, extending the Fable and Mythos lines that first launched together on July 1, 2026. They are the first confirmed model releases to follow the claude-marshmallow-eap and claude-melon-eap leak, though Anthropic has not stated whether they are the same systems the leaked identifiers referred to.
What is Model 2?
Model 2 is the internal name Anthropic used in its August 2026 risk report for a system it says outperforms Claude Mythos 5 on its internal CoBench v2 benchmark, scoring 62.8% versus 50.3% for Mythos 5. Anthropic has stated it has no current plans to release Model 2 externally.
Will Model 2 become Claude Mythos 6?
That is speculation reported by Seeking Alpha, not a confirmed naming decision from Anthropic. The company has not stated that Model 2 will ever ship publicly, under that name or any other. The September 1 release of Fable 5.1 and Mythos 5.1 was a point-release update to the existing Mythos line, not the larger Model 2-driven jump this question refers to.
What is the “Honeycomb” leak cycle mentioned in this story?
Honeycomb is the name a roadmap-tracking site used for a set of leaked codenames earlier in 2026 that correctly predicted the launches of Claude Sonnet 5, Fable 5, Mythos 5, and Opus 5 within weeks of the leaks appearing.
How does this connect to Anthropic’s reported IPO plans?
Reports of Anthropic exploring a roughly $100 billion valuation surfaced around the same time as the Model 2 disclosure. A frontier lab preparing for a major fundraising step has an incentive to show investors it has capability in reserve beyond its current public models, which is part of why the timing of these disclosures has drawn extra scrutiny.
Did Anthropic pause external security testing before this release?
Yes. Reuters reported on August 31, 2026, that Anthropic had resumed external cybersecurity testing after a pause tied to July incidents involving Claude Opus 4.7, Claude Mythos 5, and an internal research test model. That resumption came one day before Anthropic released Fable 5.1 and Mythos 5.1, though Anthropic has not said whether the internal test model involved was Model 2 or a separate system.
Should developers build against the leaked model identifiers now?
The claude-marshmallow-eap and claude-melon-eap identifiers specifically, no. Neither has a documented, stable API endpoint, early access strings can be rotated or removed without notice, and Anthropic has not published integration guidance for either name. Claude Fable 5.1 and Claude Mythos 5.1, which shipped September 1, are a different matter since they are official releases, but developers should still confirm current specs on Anthropic’s own system cards page before building against them.
When will Anthropic officially confirm what’s next?
Part of this has already happened: Fable 5.1 and Mythos 5.1 shipped on September 1, 2026. What remains unconfirmed is Model 2’s release timeline, which Anthropic has not set, and whether the marshmallow and melon identifiers map onto the models that already shipped or to something still unreleased.
