PodcastIntelligence Snacks · Weekly conversations about AI, software and business, from the team behind Normie Mode.Listen →
Models · Anthropic

Claude Haiku 4.5

Anthropic's Claude model, released 15 Oct 2025.

Top three for

No task yet.

Evidence

Independently tested. 17 independent results so far.

Cost

$5.50 per 1,000 typical requests, if you use it directly.

Reads at once

300 pages of about 500 words.

Where you can use it

The apps and plans that include it, cheapest first.

App and planPriceAccess
Claude FreeFreeIncluded
Claude Pro$20/month
$17 a month if paid yearly
Included
Claude Max$100/month
From $100 a month; a $200 tier has four times the usage
Included

Where it ranks, task by task

Its place on each task's score, among the models we compare that have enough results to score.

TaskPositionTask score
Best AI for research11th of 115
Best AI for studying and homework12th of 122
Best AI for maths and science12th of 120
Best AI for documents, PDFs and notes13th of 130
Best AI for planning and personal admin14th of 140
Best AI for translation and languages14th of 140
Best AI for writing14th of 140
Best AI for health questions14th of 140
Best AI for CVs, job applications and interviews14th of 140

Test results

Every test we use that has a result for this exact version.

TestResultPosition
People's votes on expert questions
LMArena
rating 144313th of 14
People's votes on business and finance questions
LMArena
rating 140014th of 14
People's votes on coding questions
LMArena
rating 145414th of 14
People's votes on back-and-forth conversations
LMArena
rating 140914th of 14
People's votes on creative writing
LMArena
rating 137114th of 14
People's votes on following instructions
LMArena
rating 141414th of 14
People's votes on legal questions
LMArena
rating 140014th of 14
People's votes on long, detailed requests
LMArena
rating 142714th of 14
People's votes on maths questions
LMArena
rating 139513th of 13
People's votes on health questions
LMArena
rating 139213th of 13
People's votes on questions in other languages
LMArena
rating 137614th of 14
People's votes overall
LMArena
rating 139614th of 14
People's votes on science questions
LMArena
rating 140514th of 14
People's votes on writing and language
LMArena
rating 138714th of 14
Pulling information out of documents into tables
DTBench authors
74%13th of 13
Graduate-level science questions
NYU and others
60%11th of 11
Competition maths
Epoch AI
36%12th of 12
Answering short factual questions correctly
Google DeepMind
13%11th of 11
Getting jobs done on a computer's command line
Terminal-Bench team
36%2nd of 2

Details

Maker
Anthropic
Released
15 Oct 2025
Status
Current
Name checked
Not yet confirmed with the maker
Intelligence Snacks newsletter

The big AI ideas each week, in your inbox