Claude 3 (Anthropic)
Anthropic launches the Claude 3 family (Haiku, Sonnet, Opus), with leading performance in reasoning, vision and multilingual processing.
Key points
- Anthropic launches the Claude 3 family on March 4, 2024, in three versions: Haiku, Sonnet and Opus.
- Leading performance in reasoning, vision and multilingual processing.
- Claude 3 Opus positions itself as a direct competitor to GPT-4 on many benchmarks.
- It multiplies the answer engines where a brand can be cited.
Analysis
With Claude 3, the market moves from a single dominant model to a genuine competitive landscape of high-performing assistants. Each new engine capable of answering and citing sources becomes an additional visibility channel, with its own training data and its own ways of picking up information.
For an AEO strategy, this fragmentation means you can no longer just optimize for a single assistant. The consistency of brand information across the web becomes decisive, because the different models draw from varied sources and may diverge in what they assert.
What to do
- Test your visibility across several assistants (including Claude) and compare the descriptions obtained to spot inconsistencies.
- Standardize key brand information (offering, figures, positioning) across all web presence points.
- Prioritize mentions and citations on reliable and frequently reused sources, which feed the training and answers of the models.
A serious competitor to GPT-4. Multiplies the answer engines where a brand can be cited.