Enterprise DNAEnterprise DNA

Side-by-side comparison · models

Claude 3.5 Haiku vs Claude 3 Haiku

What we have checked about each, line by line. Every value shows where it came from and when we last confirmed it.

Fast Claude model for responsive assistance, classification, and lightweight agents.

12 checked values

Claude 3 Haiku is Anthropic's fastest and most compact model for near-instant responsiveness. Quick and accurate targeted performance.

14 checked values

Checked facts

Line by line

Claude 3.5 Haiku vs Claude 3 Haiku, compared on checked facts
MeasureClaude 3.5 HaikuClaude 3 Haiku
Tokens served on OpenRouter in a dayNot checked yet3.9M (for 26 Sept 2026)Checked OpenRouter model activity (715 of 715 series)
Tool calls on OpenRouter in a dayNot checked yet81 (for 26 Sept 2026)Checked OpenRouter model activity (715 of 715 series)
Gpqa diamond38.1% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)36.3% (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)
Aider polyglot28% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)Not checked yet
Balrog19.3% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)Not checked yet
Cadeval32% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)12% (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)
Dtbench56.7% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)50.1% (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)
Frontiermath 2025 02 28 private0.3% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)Not checked yet
Geobench34% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)Not checked yet
Lech mazur writing73.5% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)Not checked yet
Math level 546.4% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)14.9% (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)
Mmlu74.3% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)73.8% (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)
Weirdml30.7% (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)9.8% (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)
Otis mock aime 2024 20254.3% (saturated benchmark: the top models all score near the ceiling) (for 22 Oct 2024)Checked Epoch AI (CC BY 4.0)1.8% (saturated benchmark: the top models all score near the ceiling) (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)
Artificial Analysis Intelligence IndexNot checked yet5.6Checked Artificial Analysis via OpenRouter
LMArena text rating (Elo)Not checked yet1,195Checked LMArena (CC-BY-4.0), published 2026-10-22
LMArena vision rating (Elo)Not checked yet950Checked LMArena (CC-BY-4.0), published 2026-10-22
LmcaNot checked yet8.8% (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)
WinograndeNot checked yet74.2% (for 7 Mar 2024)Checked Epoch AI (CC BY 4.0)

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.