Claude Fable 5.2 is now in the model picker on your CRHQ server. It is Anthropic's model for the most demanding work: agent sessions that run for hours, deep multistep research, and analysis carried through to a finished document, spreadsheet or deck. Anthropic says it writes up to 30% faster than Fable 5.1, and it costs 20% less.
Stronger across the board
These are Anthropic's reported results. The chart compares Fable 5.2 with the Fable it follows and with Opus 5.5, Anthropic's everyday recommendation.
| Benchmark | What it measures | Fable 5.2 | Fable 5.1 | Opus 5.5 |
|---|---|---|---|---|
| FrontierCode 2.0 | Agentic coding | 49.1% (Xhigh effort) | 32.4% (Xhigh effort) | 45.5% |
| FrontierSWE v2 | Agentic coding | 69.9% | 56.3% | 62.3% |
| AutomationBench | Business workflows | 51.4% | 31.4% | 42.5% |
| Terminal-Bench-Science 0.1 | Agentic scientific research | 74.6% | 52.6% | 58.7% |
| WANDR (relaxed score) | Agentic search | 82.1% | 68.7% | 72.3% |
| OSWorld 2.1 (offline set, partial) | Computer use | 89.3% | 83.5% | 87.2% |
| RoboDojo Sim | Robotics | 64.9% | 27.6% | 39.4% |
| AA-Briefcase v1.1 (rubric) | Knowledge work | 66.1% | 61.5% | 61.2% |
| AA-Briefcase v1.1 (Elo) | Knowledge work | 1785 | 1669 | 1800 |
| Remote Labor Index | Knowledge work | 41.3% (High effort) | 17.9% | 21.3% |
| Humanity's Last Exam (with tools) | Multidisciplinary reasoning | 74.2% | 65.6% | 67.7% |
| Chartography (no tools) | Reading charts and figures | 73.4% | 44.8% | 64.4% |
The biggest jumps are on long, open-ended work. On the Remote Labor Index, which uses real remote-work projects, Fable 5.2 more than doubles Fable 5.1's score. On AutomationBench and Terminal-Bench-Science it gains about 20 points. On FrontierCode 2.0, Anthropic reports that Fable 5.2 at High effort scores about level with Opus 5.5 at its highest effort, for roughly half the cost per task.
Anthropic also reports about 55% fewer unnecessary safety blocks than with Fable 5.1, so legitimate work is declined less often.
When to use it
Anthropic still recommends Opus 5.5 as the everyday model for most coding and knowledge work. Fable 5.2 is the model to move up to when a job is long, open-ended and hard: a multi-hour agent session, a deep investigation, or a task where Opus 5.5 at high effort still falls short. It can also lead a team of agents, handing simpler steps to faster models.
Pricing
| Per 1M tokens | Fable 5.2 | Fable 5.1 | Opus 5.5 |
|---|---|---|---|
| Input | $8.00 | $10.00 | $4.00 |
| Output | $40.00 | $50.00 | $20.00 |
| Cache reads | $0.20 | $0.25 | $0.20 |
| Cache writes (5 minutes) | $10.00 | $12.50 | $5.00 |
| Cache writes (1 hour) | $16.00 | $20.00 | $8.00 |
Anthropic gives Fable 5.2 a 1 million token context window and up to 128,000 tokens in one reply. Thinking is always on, and the effort setting controls how hard it works.
What changes on your CRHQ server
- Fable 5.2 is in every model picker. It is listed above Fable 5.1 under the Claude provider. Pick it for a session, or set it as an agent's model.
- Fable (latest) now means Fable 5.2. Agents and sessions set to Fable (latest) move to Fable 5.2 on their own once your server has the update.
- Fable 5.1 and Fable 5 stay available. Anything set to one of them keeps running on that exact version.
- Effort and context work as with Fable 5.1. All five effort levels are there, and CRHQ does not limit the context window.
- Claude Code is updated. Your server runs Claude Code 2.1.322, which brings Fable 5.2.
- Costs use Anthropic's prices. Each session's Costs tab tracks Fable 5.2 at the official rates above.
- Claude subscriptions work as before. Subscription rotation and the weekly Fable limit apply to Fable 5.2 the same way, with nothing to change. If a Claude login expires, CRHQ still switches to the next account.
Availability
| Fact | Value |
|---|---|
| Availability | Every CRHQ server |
| Live since | 2026-11-10 |
| Pricing | $8 in / $40 out per 1M tokens; cache reads $0.20 |
| Client action required | None. Pick Fable 5.2 in the model picker |