Glean adds Gemini 4 Carbon, Claude Fable 5.2 and Gemini 4 Hydrogen

Glean added three models: Gemini 4 Carbon, Claude Fable 5.2 and Gemini 4 Hydrogen.

Glean describes Carbon as its leader in task completion for complex workflows, Fable 5.2 as more accurate and more token efficient than Fable 5.1 on complex tasks, and Hydrogen as a balanced model for research and comprehensive answers.

Those are Glean's own assessments. Its charts show Fable 5.2 using 5.3% fewer tokens per query than Fable 5.1, and Gemini 4 Hydrogen preferred 69.47% of the time over Gemini 3.8 Flash in pairwise comparisons.