TL;DR
Anthropic is restoring Claude Fable 5 after the Commerce Department lifted export controls on June 30, with limited weekly access through July 7. OpenAI’s GPT-5.6 remains in a restricted preview for government-vetted partners, while a reported Anthropic successor model remains unconfirmed.
Anthropic is restoring Claude Fable 5 to users on July 1 after an 18-day access halt tied to U.S. export controls, putting one of the summer’s most closely watched AI models back online as OpenAI’s GPT-5.6 remains limited to government-vetted preview partners.
The restoration follows the Commerce Department’s June 30 lifting of export controls that had kept Claude Fable 5 offline for many users. According to the source material, Anthropic is bringing the model back on the Claude platform, Claude.ai, and Claude Code, with cloud availability on AWS, Google Cloud, and Microsoft Foundry expected to follow.
Anthropic is offering Fable 5 for up to 50% of weekly usage limits through July 7 on Pro, Max, Team, and selected enterprise plans. The return appears to come with tighter access rules: Anthropic has agreed to work with the government on release protocols, report malicious activity, and add safeguards aimed at blocking a specific jailbreak that officials had raised concerns about.
The company’s more restricted Mythos 5 model, described in the source material as a higher-capability sibling with fewer guardrails, has also been restored to a limited group of U.S. organizations through the Glasswing program. That access is said to cover more than 100 institutions focused on critical infrastructure defense, though the full eligibility criteria and schedule for wider access are not confirmed.
Fable 5 is back. GPT-5.6 is next. And Anthropic reportedly already has something stronger.
The most-wanted model of the summer is online again — and it may already be the second-best model Anthropic has, behind one the public has never seen. The AI you’re allowed to use is now a curated slice of the AI that exists.
Restored on Claude platform, Claude.ai & Code. Up to 50% of weekly limits through July 7. Was briefly the benchmark king — now returns with new safeguards & possible ID checks.
Previewed June 26 to only ~20 government-vetted partners; general release “in coming weeks,” pending Washington’s nod. Cheaper than Fable — roughly half the price.
OpenAI · compute-heavy
OpenAI · flagship
the tie — “Fable-5 level”
Anthropic · GA fallback
On June 21, ~9 days into the blackout, AI analyst Andrew Curran said on X that Anthropic had already finished training a more capable Mythos successor — possibly shipping as Mythos 5.1 / 6, possibly staying internal. Anthropic hasn’t confirmed it. But it’s not baseless: an unreleased Mythos Preview already sits above the public tier — OpenAI even benchmarks Sol against it. The pattern is real even if the specific model isn’t proven.
Stack it up and the shape is clear: what the public can use — Fable 5 today, GPT-5.6 in weeks, whatever clears the gate next — is a permissioned, curated slice of what these labs have actually built. A stronger tier is almost always one step ahead, behind a government gate or a lab’s caution — and both companies are pushing to make that review process permanent. For builders the instruction is blunt: don’t chase “the best model.” Build so you can swap whichever one you’re allowed to use this week — because that list keeps changing.
Model Access Is Being Filtered
The immediate impact is practical: developers, companies, and researchers who had planned around Fable 5 can use it again, but not necessarily with the same access they expected before the halt. The restoration shows that frontier model availability is now shaped not only by lab release plans, but also by government review, usage limits, and identity controls.
That matters because OpenAI’s GPT-5.6 is in a similar position. OpenAI previewed the model family on June 26 to about 20 government-vetted partners, according to the source material, with broader release described as coming in the weeks ahead. The company has said it does not want this kind of gated release process to become normal, but for now the pattern is clear: users may be waiting on permission as much as product readiness.
The GPT-4 Millionaire: Future of Business Featuring Microsoft 365 Copilot: How to Leverage AI Language Models to Grow Your Company and How AI-driven Language Models Will Revolutionize the Way We Work
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Benchmarks Show A Split Race
Before it went offline, Fable 5 was described as one of the strongest commercially available models, with high marks on coding tasks and strong early user reports. The source material says it briefly led coding leaderboards and cites early testing in which Stripe reportedly used it to overhaul a 50-million-line codebase in a single day, though that result should be treated as a reported example rather than independently verified performance data.
OpenAI’s GPT-5.6 preview is presented as a three-tier family: Sol, Terra, and Luna, with a higher-compute Sol Ultra mode. On the vendor preview numbers cited in the source material, GPT-5.6 Terra ties Fable 5 on Terminal-Bench 2.1 at 84.3, while GPT-5.6 Sol scores 88.8 and Sol Ultra scores 91.9. Those figures come from preview materials and have not been independently verified.
AI-Assisted Programming: Better Planning, Coding, Testing, and Deployment
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Stronger Anthropic Model Unproven
The largest unresolved claim is whether Anthropic has already trained a more capable successor to Mythos that remains unreleased. The source material attributes that claim to AI analyst Andrew Curran, who said on X on June 21 that Anthropic had completed such a model, possibly under a Mythos 5.1 or Mythos 6 name. Anthropic has not confirmed that claim.
There is also uncertainty around Fable 5 access terms. It is not yet clear whether all subscribers regain the earlier level of access, whether some usage will sit behind credits, or how identity checks will apply across plans and regions. For GPT-5.6, the open question is when Washington will allow broader access and whether the final release will match the previewed capability and pricing claims.
cloud-based AI model hosting services
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Release Gates Now Set Pace
The next milestones are Anthropic’s completion of Fable 5 re-enablement across cloud platforms and OpenAI’s move from GPT-5.6 preview to general availability. Readers should watch for updated access terms, pricing, independent benchmark testing, and any formal Anthropic statement on the reported Mythos successor.
For builders, the near-term lesson is that the highest-performing model may change quickly, and access may depend on policy review as much as technical progress. Systems that can switch between Fable 5, GPT-5.6, and approved fallback models will be better placed than workflows tied to a single provider.
Agentic AI: The Practitioner's Guide: Build Production-Ready AI Agents with LangChain, LangGraph, RAG, MCP, and FastAPI
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Is Claude Fable 5 available again?
Yes, Anthropic began restoring Claude Fable 5 on July 1, 2026, after the Commerce Department lifted export controls on June 30. Access is returning first through Anthropic’s own Claude services, with cloud platforms expected to follow.
Is GPT-5.6 publicly available?
Not broadly. According to the source material, OpenAI previewed GPT-5.6 on June 26 to about 20 government-vetted partners. General release has been described as coming in the weeks ahead, but no confirmed date is provided.
Is GPT-5.6 better than Fable 5?
The answer depends on the tier and the benchmark. The cited preview numbers show GPT-5.6 Terra tying Fable 5 on Terminal-Bench 2.1, while Sol and Sol Ultra score higher. Those are vendor preview figures, not independent results.
Has Anthropic confirmed a stronger Mythos successor?
No. The claim that Anthropic has trained a stronger Mythos successor is attributed to Andrew Curran on X and remains unconfirmed. The existence of unreleased internal models is plausible, but this specific model has not been publicly verified.
Source: Thorsten Meyer AI
