Side-by-side comparison · models
Claude 3.5 Haiku vs Claude Fable 5.1 (batch)
What we have checked about each, line by line. Every value shows where it came from and when we last confirmed it.
Models
Claude 3.5 Haiku
Fast Claude model for responsive assistance, classification, and lightweight agents.
Models
Claude Fable 5.1 (batch)
Claude Fable 5.1 improves on Claude Fable 5 across the board, with the biggest gains in agentic coding, long-running agentic workflows, and knowledge work: long code refactors.
Checked facts
Line by line
| Measure | Claude 3.5 Haiku | Claude Fable 5.1 (batch) |
|---|---|---|
| Latest version | Not checked yet | anthropic/claude-fable-5.1-20260831 |
| Time to first token (ms) | Not checked yet | 97,890 |
| Output speed (tokens a second) | Not checked yet | 61.3 |
| Gpqa diamond | 38.1% (for 22 Oct 2024) | Not checked yet |
| Aider polyglot | 28% (for 22 Oct 2024) | Not checked yet |
| Balrog | 19.3% (for 22 Oct 2024) | Not checked yet |
| Cadeval | 32% (for 22 Oct 2024) | Not checked yet |
| Dtbench | 56.7% (for 22 Oct 2024) | Not checked yet |
| Frontiermath 2025 02 28 private | 0.3% (for 22 Oct 2024) | Not checked yet |
| Geobench | 34% (for 22 Oct 2024) | Not checked yet |
| Lech mazur writing | 73.5% (for 22 Oct 2024) | Not checked yet |
| Math level 5 | 46.4% (for 22 Oct 2024) | Not checked yet |
| Mmlu | 74.3% (for 22 Oct 2024) | Not checked yet |
| Weirdml | 30.7% (for 22 Oct 2024) | Not checked yet |
| Otis mock aime 2024 2025 | 4.3% (saturated benchmark: the top models all score near the ceiling) (for 22 Oct 2024) | Not checked yet |
A line appears only when at least one record has a value we confirmed from a source. Values we have not confirmed are left out, never guessed. Records of different types share fewer lines.
Add a third
Compare with one more record
How this works
Share it, or build your own
This page works for any two or three records in the directory, of any type. The link stays the same whichever order you name them in, so you can share it as it is. Change the records in the address, for example /directories/vs/claude-code-vs-cursor, or start from our editorial comparisons.