Claude Fable 5.2
Knows up to June 2026Missing 1,712 events
claude-fable-5-2
$8 in, $40 out per 1M tokens
Claude Mythos 5.2
The unrestricted variant of Fable 5.2 for defensive cybersecurity and biomedical research.
Model idclaude-mythos-5-2
Price$8 in, $40 out per 1M tokens
Checked9 November 2026
Knows up to June 2026
1,712events after its cutoff
79 historic, 388 major. 132 days of news.
It launched on 9 November 2026, already missing 1,709 of them.
How to call it
2 ways to use Claude Mythos 5.2, with the model id each provider expects.
| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Anthropic (Cyber Verification Program / Life Sciences Verification Program) | claude-mythos-5-2 |
anthropic.com | |
| Claude Security (Claude Enterprise) | claude.ai |
Pricing
| Input | $8 per 1M tokens |
|---|---|
| Output | $40 per 1M tokens |
As published: “per 1M tokens (USD); fast mode $16 / $80 (Verification Program members only)”.
Source: anthropic.com, checked 9 November 2026.
Specs
- Input
- Text, images and PDF
- Output
- Text
- Context window
- 1,000,000 tokens
- Max output
- 128,000 tokens
- Open weights
- No
- Licence
- Proprietary
- Training cutoff
- June 2026
- Released
- 9 November 2026
- Status
- Current
- Type
- Reasoning model
What stands out
What its maker marketed at launch, and what people found later, each with its source.
Vulnerability discovery
CVE Rediscovery (1,829 published CVEs): 89.6%, 10.2% more vulnerabilities than Opus 5.5 and 14.4% more than Mythos 5.1; Mozilla says it roughly doubled Mythos 5.1’s discovery metrics on Firefox. CyScenarioBench 76.3%.
Whole-capsid structure prediction
Wrote a method on top of OpenFold3-preview2 and ESMFold2 that predicts whole AAV capsid structures from overlapping chain groups plus known symmetry; 28 of 34 capsids within 3 Å in Anthropic’s tests.
Lower sandbox-escape propensity
In a 1,800-scenario adversarial sandbox evaluation it used planted out-of-sandbox credentials in 3.1% of scenarios, versus 21% for Mythos 5.1 and 39% for Mythos 5; automated behavioral audit score 2.67 (lower is better) vs 3.28 for Mythos 5.1.
Four assistants, one scale
Claude Mythos 5.2 is added as a fifth row. Each row runs from what the model knows (graphite) through the months between its cutoff and its launch (hatched) to today (pink).
- GPT-6 Astra 271 at launch, 1,780 today
- Claude Opus 5.5 372 at launch, 1,712 today
- Gemini 4 Argon 1,121 at launch, 1,806 today
- Grok 4.7 387 at launch, 1,746 today
- Claude Mythos 5.2 1,709 at launch, 1,712 today
Show as a table
| Model | Cutoff | Released | Missing at launch | Missing today |
|---|---|---|---|---|
| GPT-6 Astra | April 2026 | 3 September 2026 | 271 | 1,780 |
| Claude Opus 5.5 | June 2026 | 22 September 2026 | 372 | 1,712 |
| Gemini 4 Argon | March 2026 | 13 October 2026 | 1,121 | 1,806 |
| Grok 4.7 | May 2026 | 21 September 2026 | 387 | 1,746 |
| Claude Mythos 5.2 | June 2026 | 9 November 2026 | 1,709 | 1,712 |
Notes
Same underlying model as Claude Fable 5.2 without its cybersecurity and biology classifiers; access only for vetted organizations through Anthropic’s Cyber and Life Sciences Verification Programs, and inside Claude Security for Enterprise customers.