Claude Mythos 5.1
Claude Mythos 5.1 is a verified current AI model with official specifications, pricing or access details, capabilities, practical use cases, strengths and limitations.
What is this model and why does it matter?
Claude Mythos 5.1 is Anthropic's September 2026 specialized Project Glasswing model, sharing Fable 5.1-class capabilities for demanding long-horizon reasoning while using always-on adaptive thinking.
Claude Mythos 5.1: features, use cases and important details
Claude Mythos 5.1 is a current AI model verified from first-party Anthropic sources.
Claude Mythos 5.1 verified specifications
Claude Mythos 5.1 is Anthropic’s September 2026 specialized Project Glasswing model, sharing Fable 5.1-class capabilities for demanding long-horizon reasoning while using always-on adaptive thinking. Its verified context or usage limit is 1000000 tokens, with 128000 tokens maximum output.
Claude Mythos 5.1 pricing and access
Claude Mythos 5.1 costs $10 per 1M input tokens and $50 per 1M output tokens. Cache reads cost $0.25/M tokens and Batch API input/output receives a 50% discount. Access is invitation-only.
Claude Mythos 5.1 best uses
Advanced research, Cybersecurity research, Biology research, Long-horizon agents, Complex reasoning.
Claude Mythos 5.1 limitations
Not generally available, Always-on thinking can increase output usage, Access depends on Project Glasswing eligibility.
Claude Mythos 5.1 capabilities and use cases
In addition, its main capabilities include 1M context, 128K output, Always-on adaptive thinking, Vision, Tool use and Structured outputs. For example, common use cases include Specialized research, Long-running agents, Deep analysis and High-complexity knowledge work.
Who should consider Claude Mythos 5.1?
In practice, this model may suit Advanced research, Cybersecurity research, Biology research, Long-horizon agents and Complex reasoning. Also, notable strengths include Newest specialized Claude model, 1M context, 128K output and Strong long-horizon reasoning. However, review trade-offs such as Not generally available, Always-on thinking can increase output usage and Access depends on Project Glasswing eligibility before adopting it.
Claude Mythos 5.1 pricing and access
Meanwhile, Claude Mythos 5.1 costs $10 per 1M input tokens and $50 per 1M output tokens. Cache reads cost $0.25/M tokens and Batch API input/output receives a 50% discount. Access is invitation-only. Premium $10/M input and $50/M output pricing; access requires invitation through Project Glasswing.
Official resources and verification
Use the official model website, official documentation, pricing or release source and additional primary source to confirm current availability, limits and pricing. Product details can change after publication, so rely on primary documentation for final decisions.
Compare with other AI models
Next, continue your research in the AI models directory, Anthropic models and Specialized Reasoning Model models. Compare providers, pricing, modalities and practical limitations side by side to choose the right model for your workflow.
How to use this model
- Obtain Project Glasswing access.
- Create Claude API or supported cloud credentials.
- Call claude-mythos-5-1.
- Provide text or image context.
- Use tools and structured outputs as needed.
- Monitor long-running reasoning cost.
Example prompts
Analyze this complex research problem and identify the strongest evidence.Plan this multistep expert workflow.Review these technical documents and surface high-confidence findings.
What it can do
- 1M context
- 128K output
- Always-on adaptive thinking
- Vision
- Tool use
- Structured outputs
- Low cache-read price
Practical use cases
- Specialized research
- Long-running agents
- Deep analysis
- High-complexity knowledge work
What does it cost?
Claude Mythos 5.1 costs $10 per 1M input tokens and $50 per 1M output tokens. Cache reads cost $0.25/M tokens and Batch API input/output receives a 50% discount. Access is invitation-only.
What stands out
- Newest specialized Claude model
- 1M context
- 128K output
- Strong long-horizon reasoning
Things to consider
- Invite-only
- High token cost
- Proprietary
Important restrictions and trade-offs
- Not generally available
- Always-on thinking can increase output usage
- Access depends on Project Glasswing eligibility
Our editorial take
A very fresh September 2026 Claude model with strong high-intent interest around Anthropic’s most specialized research tier.