Claude Fable 5.2 vs Claude Opus 5
Claude Fable 5.2 leads Claude Opus 5 by 11.0 points on the Shogo Agent Index. Claude Fable 5.2 wins 19 and Claude Opus 5 wins 1 of 20 shared benchmarks. Claude Opus 5 is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | Claude Fable 5.2 | Claude Opus 5 |
|---|---|---|
| Agentic coding | 86 | 74 |
| Long-horizon autonomy | 78 | 74 |
| Knowledge-work agents | 67 | 52 |
| Computer use | 77 | 56 |
| Security & context learning | 65 | 59 |
Shared benchmarks
| Benchmark | Claude Fable 5.2 | Claude Opus 5 |
|---|---|---|
| APEX-Agentsindex | 80.1% | 65.8% |
| Vending-Bench 2index | $9,628 | $11,182 |
| FrontierSWEindex | 69.9% | 52.0% |
| CursorBenchindex | 60.0% | 46.6% |
| PostTrainBenchindex | 44.6% | 35.7% |
| ARC-AGI | 98.5% | 97.5% |
| ARC-AGI-2 | 93.8% | 90.4% |
| Chess Puzzles | 79.0% | 43.0% |
| CritPt | 35.7% | 29.1% |
| DTBench | 99.2% | 97.9% |
| FrontierMath Tier 4 | 100.0% | 73.2% |
| FrontierMath Tiers 1-3 | 95.1% | 85.6% |
| GPQA Diamond | 96.0% | 93.9% |
| LMCA | 70.6% | 64.5% |
| OTIS Mock AIME | 99.7% | 98.9% |
| ProofBench | 100.0% | 99.0% |
| SciCode | 67.7% | 56.4% |
| SimpleBench | 94.9% | 80.6% |
| SimpleQA Verified | 94.0% | 59.9% |
| WeirdML v3 | 43.1% | 19.0% |
Specs and price
| Claude Fable 5.2 | Claude Opus 5 | |
|---|---|---|
| Agent Index | 76.4 (#1) | 65.4 (#11) |
| Agent $/M | $3.21 | $2.31 |
| Input / output $/M | $8.00 / $40.00 | $5.00 / $25.00 |
| Context | 1M | 1M |
| Released | Nov 9, 2026 | Jul 24, 2026 |
| Weights | Closed | Closed |
| API hosts | 4 | 6 |
Don’t pick one. Route.
Shogo can run Claude Fable 5.2 and Claude Opus 5 side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo