Understand and compare
Claude 2
vs.
Claude 3 Haiku
Try
Podial
Turn your documents into engaging podcast discussions.
Overview
Claude 2 was released
8 months before
Claude 3 Haiku.
Claude 2
|
Claude 3 Haiku
|
|
|---|---|---|
|
Provider
The entity that provides this model.
|
Anthropic
|
Anthropic
|
|
Input Context Window
The number of tokens supported by the input context window.
|
100K
tokens
|
200K
tokens
|
|
Maximum Output Tokens
The number of tokens that can be generated by the model in a single request.
|
Not specified.
|
4,096
tokens
|
|
Release Date
When the model was first released.
|
2023-07-11
|
2024-03-13
|
|
Knowledge Cutoff
Limit on the knowledge base used by the model.
|
Early 2023
|
August 2023
|
|
Open Source
|
|
|
|
API Providers
The providers that offer this model. (This is not an exhaustive list.)
|
|
|
Pricing
Claude 2 is
roughly 32x more expensive compared
to Claude 3 Haiku for input tokens and
roughly 19x more expensive
for output tokens.
|
Claude 2
|
Claude 3 Haiku
|
|
|---|---|---|
|
Input
Cost of input data provided to the model.
|
$8.00
per million tokens
|
$0.25
per million tokens
|
|
Output
Cost of output tokens generated by the model.
|
$24.00
per million tokens
|
$1.25
per million tokens
|
Benchmarks
Compare relevant benchmarks between Claude 2
and Claude 3 Haiku.
|
Claude 2
|
Claude 3 Haiku
|
|
|---|---|---|
|
MMLU
Evaluating LLM knowledge acquisition in zero-shot and few-shot settings.
|
78.5
(5-shot)
|
76.7
(5-shot CoT)
|
|
MMMU
A wide ranging multi-discipline and multimodal benchmark.
|
Benchmark not available.
|
50.2
|
|
HellaSwag
A challenging sentence completion benchmark.
|
Benchmark not available.
|
85.9
(10-shot)
|
|
GSM8K
Grade-school math problems benchmark.
|
Benchmark not available.
|
Benchmark not available.
|
|
HumanEval
A benchmark to measure functional correctness for synthesizing programs from docstrings.
|
Benchmark not available.
|
Benchmark not available.
|
|
MATH
Benchmark performance on Math problems ranging across 5 levels of difficulty and 7 sub-disciplines.
|
Benchmark not available.
|
Benchmark not available.
|
Claude 2, developed by Anthropic, features a large context window of 100,000 tokens. The model costs 0.8 cents per thousand tokens for input and 2.4 cents per thousand tokens for output. It was released on July 11, 2023, and has shown strong performance in the MMLU benchmark with a score of 78.5 in a 5-shot scenario.
Claude 3 Haiku, developed by Anthropic, features a context window of 200,000 tokens. The model costs 0.025 cents per thousand tokens for input and 0.125 cents per thousand tokens for output. It was released on March 13, 2024. In benchmarks, it achieved a score of 50.2 in MMMU, 85.9 in HellaSwag in a 10-shot scenario, and 76.7 in MMLU in a 5-shot CoT scenario.
GPT-3.5 Turbo 1106
Gemini Flash
Measure & Improve LLM
Product Performance.
Get Started
© 2023-2024 Context