When was Claude 3 Haiku released?

Claude 3 Haiku was released by Anthropic on Monday, Mar 4 2024.

Who made Claude 3 Haiku?

Claude 3 Haiku was built by Anthropic. AI safety company building the Claude family of models. Founded in 2021 by former OpenAI researchers.

What benchmark scores did Claude 3 Haiku get?

Claude 3 Haiku reports 2 tracked benchmark scores — BullshitBench v2: 10%; GPQA Diamond: 33.3%. Tracked scores may come from lab reports or independent benchmarks; source details accompany the benchmark data.

What is the context window of Claude 3 Haiku?

Claude 3 Haiku has a context window of 200k. That is the maximum amount of input plus output the model can hold in a single request.

Is Claude 3 Haiku open source?

No. Claude 3 Haiku is a closed model. The weights are not published — it is available only through the provider's own API, apps, or partner platforms.

What came before and after Claude 3 Haiku?

Anthropic's previous tracked release was Claude 2.1 on Nov 21 2023, 104 days earlier. It was followed by Claude 3 Sonnet on Mar 4 2024.

Claude 3 Haiku

Released Mar 4 2024

Claude 3 Haiku is an AI model released by Anthropic on Monday, Mar 4 2024, 104 days after Claude 2.1. It has a 200k token context window. Benchmark results (shown below) cover BullshitBench v2, and GPQA Diamond.

Benchmarks

Reasoning & science

GPQA Diamond

33.3%

#66 of 66

Robustness

BullshitBench v2

10%

#86 of 97

Compare Claude 3 Haiku with

Claude 3 Haiku

Suggested comparisons

Claude 3 HaikuvsClaude Haiku 5.5 Claude 3 HaikuvsGPT-6.1 Sol Claude 3 HaikuvsGemini 4 Argon Claude 3 HaikuvsMuse Spark Open Claude 3 HaikuvsGrok 4.7 Claude 3 HaikuvsDeepSeek-V4.1-Flash-1017 Claude 3 HaikuvsMistral Large 4 Claude 3 HaikuvsKimi K3 Claude 3 HaikuvsGLM-5.3-Flash Claude 3 HaikuvsQwen3.8-Max-0902 Claude 3 HaikuvsNemotron 3.5 Super VL

Frequently asked questions

: Claude 3 Haiku was released by Anthropic on Monday, Mar 4 2024.
: Claude 3 Haiku was built by Anthropic. AI safety company building the Claude family of models. Founded in 2021 by former OpenAI researchers.
: Claude 3 Haiku reports 2 tracked benchmark scores — BullshitBench v2: 10%; GPQA Diamond: 33.3%. Tracked scores may come from lab reports or independent benchmarks; source details accompany the benchmark data.
: Claude 3 Haiku has a context window of 200k. That is the maximum amount of input plus output the model can hold in a single request.
: No. Claude 3 Haiku is a closed model. The weights are not published — it is available only through the provider's own API, apps, or partner platforms.
: Anthropic's previous tracked release was Claude 2.1 on Nov 21 2023, 104 days earlier. It was followed by Claude 3 Sonnet on Mar 4 2024.