RESTRICTED ACCESSMYTHOS-CLASSGLASSWINGTEXT + VISIONNO PUBLIC SIGN-UP
Claude Mythos 5 API Pricing
Claude Mythos 5 is Anthropic's restricted-access Mythos-class flagship: the same underlying model as Claude Fable 5, with safeguards lifted in some areas for approved cyberdefenders and infrastructure partners. Anthropic lists it on its public price table - $10/M input, $50/M output, $1.00/M cache hits - but access runs only through Project Glasswing.
Restricted input - per 1M tokens
$10.00/M
Source Anthropic launch restricted
Restricted output - per 1M tokens
$50.00/M
Access Glasswing / trusted restricted
Cached input - per 1M tokens
$1.00/M
Prompt cache hit price -90%
Effective - agentic blend
$6.82/M
92/8 split - 82% cache
§ 01 / TERMINAL
Run the numbers.
The rate card is public even though access is not, so this calculator runs on Anthropic's listed Mythos 5 rates, cache writes and cache hits included. It applies to an account approved through Project Glasswing.
$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
§ 02 / SCENARIOS
Real-world presets.
CYBER DEFENSE
Vulnerability triage pass
$0.726/task
INFRASTRUCTURE
Large-codebase security review
$1.09/review
LIFE SCIENCES
Protein-design tool run
$0.768/run
RESEARCH
Novel hypothesis work session
$0.338/session
Estimate · anthropic-bpe-estimate · ≈2.6 chars/token Auto-counts as you type
This is a chars-per-token approximation, not a real tokenizer. Actual tokens vary by language, code density, and tool-call overhead — counts are typically ±10–20% off for English prose, more for code or non-Latin scripts. For exact billing, use the vendor's official tokenizer.
Characters 448
Words 63
Tokens (estimated) 172 tokens
Cost as input · uncached $0.00172 USD
Cost as output · uncached $0.0086 USD
Cost as cached input $0.00017 USD
| Model | Input /M | Output /M | Effective blended | Context | Best for |
|---|---|---|---|---|---|
| Claude Mythos 5 Current | $10.00 cache $1.00 | $50.00 | $6.82 agentic 92/8 | 1M | Approved cyber / bio trusted access |
| Claude Mythos 5.1 | $10.00 cache $0.25 | $50.00 | $6.26 −8% | 1M | Newer Mythos, cheaper cache reads |
| Claude Fable 5 | $10.00 cache $1.00 | $50.00 | $6.82 same | 1M | Generally available flagship usage |
| Claude Mythos Preview | $25.00 | $125 | $33.00 withdrawn | documented elsewhere | Historical restricted preview |
| Claude Opus 4.8 | $5.00 cache $0.50 | $25.00 | $3.41 −2.0× | 1M | Public frontier fallback |
| Claude Sonnet 4.6 | $3.00 cache $0.30 | $15.00 | $2.05 −3.3× | 1M | Production agents and coding |
| GPT-5.5 | $5.00 cache $0.50 | $30.00 | $3.61 −47% | 1M | OpenAI frontier agents |
| Gemini 3.6 Flash | $0.75 cache $0.075 | $3.75 | $0.481 −14× · promo | 1M | Google multimodal agents |
| DeepSeek V4 Pro | $1.32 cache $0.044 | $3.96 | $0.569 −12× | 1M | Low-cost reasoning workloads |
Frequently asked.
Practical Claude Mythos 5 pricing questions, with access limits separated from Fable 5 general availability.
Q · 01 What is Claude Mythos 5 priced at? +
Anthropic's pricing table lists
$10/M input, $50/M output, $12.50/M 5-minute and $20/M 1-hour cache writes, $1.00/M cache hits, and batch at $5/$25. That is the same card as Claude Fable 5, and less than half the earlier Mythos Preview rate. Under this site's 92/8 agentic blend at an 82% cache-hit rate it works out to $6.82/M.Q · 02 Can anyone use Mythos 5? +
No. Anthropic marks it limited availability through Project Glasswing, for vetted organisations; there is no self-serve sign-up. The price is public anyway, which is why this page can carry it.
Q · 03 What is the API model ID for Mythos 5? +
claude-mythos-5. Anthropic's model-deprecations table lists it as active, with retirement not sooner than June 9, 2027. Calls on that ID only work for accounts approved through Project Glasswing.Q · 04 How is Mythos 5 different from Fable 5? +
Anthropic says Mythos 5 is the same underlying model as Fable 5, but with safeguards lifted in some areas for approved users. Fable 5 is the generally available version with safeguards and Opus 4.8 fallback.
Q · 05 What kinds of users get Mythos 5 access? +
Anthropic says current Mythos Preview / Project Glasswing users can upgrade first, and it plans broader trusted access for cybersecurity organizations plus a biology trusted-access program.
Q · 06 Does Mythos 5 have prompt-caching pricing? +
Yes. Anthropic's pricing table lists cache hits at
$1.00/M, a tenth of base input, with cache writes at $12.50/M for five minutes and $20/M for one hour. The calculator and the effective rate on this page use them. Claude Mythos 5.1 cut the cache hit to $0.25/M and left the other rates where they are.Q · 07 Is Mythos 5 replacing Mythos Preview? +
For existing Glasswing partners, yes in practice: Anthropic says users with Mythos Preview access will be able to upgrade to Mythos 5.
Q · 08 Should general developers use Mythos 5 or Fable 5? +
General developers should use
claude-fable-5. Mythos 5 is restricted; Fable 5 is the generally available model with safeguards.Q · 09 Is there a newer Mythos? +
Yes: Claude Mythos 5.1, released September 1, 2026 through the same Glasswing programme. Base input, cache writes, output and batch are the same as on Mythos 5; only the cache hit is cheaper. Mythos 5 stays active and on the price table.