Compare / head-to-head
Claude Fable 5.1vs
Step 5 Preview
Claude Fable 5.1 leads 5 of 5 shared benchmarks. Step 5 Preview is 16.2x cheaper per token. Both offer a 1M-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
5 – 0
Claude Fable 5.1 leads
Cheaper per token
Step 5 Preview
16.2x cheaper, input + output
Larger context
Tie
both 1M tokens
Providers
2 providers
Anthropic · StepFun
Claude Fable 5.1
Anthropic · released 2026-09-01
textvision
Step 5 Preview
StepFun · released 2026-09-20
textvisionvideo
Quality
Benchmark matrix
5 shared · 27 only Claude Fable 5.1 · 0 only Step 5 Preview| Benchmark | Claude Fable 5.1 | Step 5 Preview | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 5 | ||||
| LMArena Elo | 1501.1 ↗ | 1453.7 ↗ | +47.4 | Claude Fable 5.1 |
| LMArena Agent (LMArena) | 0.1431 ↗ | 0.0048 ↗ | +0.1 | Claude Fable 5.1 |
| LMArena Vision (LMArena) | 1287.7 ↗ | 1259.7 ↗ | +28 | Claude Fable 5.1 |
| LMArena WebDev (LMArena) | 1749.2 ↗ | 1570.2 ↗ | +179 | Claude Fable 5.1 |
| Terminal-Bench 4.0 (Vals AI) | 58.1% ↗ | 31.8% ↗ | +26.3 pt | Claude Fable 5.1 |
| Only Claude Fable 5.1 reports · 27 | ||||
| Humanity's Last Exam (no tools) | 60.9% ↗ | not reported | — | — |
| APEX-Agents (Mercor) | 68.6% ↗ | not reported | — | — |
| Chess Puzzles (Epoch AI run) | 47% ↗epoch run | not reported | — | — |
| CursorBench 3.2.0 | 73.4% ↗ | not reported | — | — |
| EBR-bench (Epoch AI run) | 57.1% ↗epoch run | not reported | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | 87.8% ↗epoch run | not reported | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 90.2% ↗epoch run | not reported | — | — |
| FrontierSWE V2 (Proximal Labs) | 56.3% ↗ | not reported | — | — |
| Furniture Assembly (Epoch AI run) | 70% ↗epoch run | not reported | — | — |
| GDPval-AA v2 | 1853 ↗ | not reported | — | — |
| GSO Opt@1 (GSO) | 87.3% ↗ | not reported | — | — |
| LiveBench Agentic Coding (LiveBench) | 66.1% ↗ | not reported | — | — |
| LiveBench Coding (LiveBench) | 86.4% ↗ | not reported | — | — |
| LiveBench Data Analysis (LiveBench) | 80.3% ↗ | not reported | — | — |
| LiveBench Instruction Following (LiveBench) | 73% ↗ | not reported | — | — |
| LiveBench Language (LiveBench) | 89.5% ↗ | not reported | — | — |
| LiveBench Mathematics (LiveBench) | 97% ↗ | not reported | — | — |
| LiveBench Reasoning (LiveBench) | 91.7% ↗ | not reported | — | — |
| MirrorCode (Epoch AI run) | 73.3% ↗epoch run | not reported | — | — |
| Mystery Game Puzzles (Epoch AI run) | 58% ↗epoch run | not reported | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 100% ↗epoch run | not reported | — | — |
| SAGE (Vals AI) | 48.5% ↗ | not reported | — | — |
| SimpleBench (SimpleBench) | 86.6% ↗ | not reported | — | — |
| SimpleQA Verified | 70.8% ↗epoch run | not reported | — | — |
| Terminal-Bench Science 0.1 | 52.6% ↗ | not reported | — | — |
| Vending-Bench 2 (Andon Labs) | 5421.56 ↗ | not reported | — | — |
| WeirdML (Håvard Tveit Ihle) | 92.9% ↗ | not reported | — | — |
| Only Step 5 Preview reports · 0 | ||||
| Step 5 Preview reports nothing Claude Fable 5.1 does not. | ||||
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Claude Fable 5.1 minus Step 5 Preview in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Claude Fable 5.1 | Step 5 Preview | Edge |
|---|---|---|---|
| Context window | 1M tokens | 1M tokens | Tie |
| Max output | 128K tokens | 64K tokens | Claude Fable 5.1 |
| Input price / 1M | $10 ↗ | $1 ↗ | Step 5 Preview |
| Output price / 1M | $50 ↗ | $2.7 ↗ | Step 5 Preview |
| Cached input / 1M | $0.25 ↗ | $0.05 ↗ | Step 5 Preview |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $60 | $3.7 | Step 5 Preview |
| Modalities | text · vision | text · vision · video | Step 5 Preview |
| Released | 2026-09-01 | 2026-09-20 | — |
| Cited benchmark scores | 36 | 5 | — |
Reliability
Provider status
Live from /statusMore matchups
Claude Fable 5.1 vs …
Models sharing the most benchmarksMore matchups
Step 5 Preview vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Claude Fable 5.1 and Step 5 Preview on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.