vibecodingnav
中/EN
Models/claude-fable-5

Claude Fable 5

Second on Android Bench 1.0 at 90.9%, run cost $144.3.

Anthropic · Proprietary · Fetched 2026-10-10

LMArena WebDev
1625
#21 / 142 · ±6
LMArena Image-to-WebDev
1622
#10 / 64 · ±9
Terminal-Bench 4.0
44.5%
#18 / 35 · ±3.8

Benchmark scores

Boards older than 90 days are not counted as recommendations
BenchmarkSettingScoreRankCostAs of
LMArena WebDevhigh1625±6#21—2026-10-08
LMArena Image-to-WebDevhigh1622±9#10—2026-10-09
Terminal-Bench 4.0max · Claude Code44.5±3.8#18$7.3k2026-10-06
The Slop Index—52#29—2026-10-11
ReactBenchmax44.642.1–47.2#4$12.762026-10-10
DeepSWE v1.1xhigh · mini-swe-agent69.966.7–73.2#5$13.412026-09-22
SWE-bench Verified—This board has stopped updating——2026-02-26
Aider Polyglot—This board has stopped updating——2025-11-20

Use it or not

Why
  • 0.9 points behind first on standard Android tasks.
When not
  • The run still costs $144.3. On a budget, look at Luna at $7.2.

Appears in these scenarios

Collected from scenario boards
Build a Web frontend#4ReactBench 44.6%Build an Android app#5DeepSWE 69.9%Write CLI, scripts, and DevOps#5DeepSWE 69.9%Write docs & copy#29Slop Index 52

Peers