Claude Fable 5.1 vs GPT-5.4
Claude Fable 5.1 leads GPT-5.4 by 16.6 points on the Shogo Agent Index. Claude Fable 5.1 wins 17 and GPT-5.4 wins 1 of 19 shared benchmarks. GPT-5.4 is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | Claude Fable 5.1 | GPT-5.4 |
|---|---|---|
| Agentic coding | 80 | 57 |
| Long-horizon autonomy | 71 | 61 |
| Knowledge-work agents | 55 | 40 |
| Computer use | 62 | 42 |
| Security & context learning | 61 | 54 |
Shared benchmarks
| Benchmark | Claude Fable 5.1 | GPT-5.4 |
|---|---|---|
| APEX-Agentsindex | 68.6% | 52.4% |
| Vending-Bench 2index | $5,422 | $6,144 |
| GSO-Benchindex | 88.2% | 31.4% |
| PostTrainBenchindex | 44.6% | 17.8% |
| ARC-AGI | 97.5% | 93.7% |
| ARC-AGI-2 | 90.0% | 74.0% |
| Chess Puzzles | 48.0% | 48.0% |
| CritPt | 31.1% | 23.4% |
| DTBench | 97.6% | 94.4% |
| FrontierMath Tier 4 | 87.8% | 49.0% |
| FrontierMath Tiers 1-3 | 90.2% | 78.6% |
| Humanity's Last Exam | 46.5% | 36.2% |
| LMCA | 65.5% | 52.0% |
| MirrorCode | 93.3% | 23.3% |
| OTIS Mock AIME | 100.0% | 97.8% |
| ProofBench | 100.0% | 56.0% |
| SciCode | 63.1% | 56.6% |
| SimpleQA Verified | 70.8% | 45.1% |
| WebDev Arena | 1758 | 1463 |
Specs and price
| Claude Fable 5.1 | GPT-5.4 | |
|---|---|---|
| Agent Index | 68.2 (#10) | 51.6 (#33) |
| Agent $/M | $4.01 | $1.27 |
| Input / output $/M | $10.00 / $50.00 | $2.50 / $15.00 |
| Context | 1M | 1.1M |
| Released | Sep 1, 2026 | Mar 5, 2026 |
| Weights | Closed | Closed |
| API hosts | 3 | 4 |
Don’t pick one. Route.
Shogo can run Claude Fable 5.1 and GPT-5.4 side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo