Today, we're excited to announce that Google’s latest flagship model, Gemini 3, is now available across the Edisource platform.
Best In Class Performance And Intelligence
Gemini 3 represents a significant leap in cross-media understanding. While previous models treated video and audio as secondary inputs, Gemini 3 processes them with the same native fidelity as text. We anticipate it will excel at the kind of heavy-lift multimedia auditing that is becoming central to modern newsrooms.
Gemini 3 has opened a massive performance lead in high-volume data processing. The most critical difference for publishers is the context window: Gemini 3 supports over 1 million tokens compared to GPT-5’s 400k limit, allowing editorial teams to analyze entire legal archives or full-length books without truncation.
When it comes to speed, Gemini 3’s native multimodal architecture processes video and audio roughly 3x faster than GPT-5 in our internal benchmarks.
While Gemini 2.5 Pro has been our go-to for logical deduction, Gemini 3 offers a different approach to reasoning that we find particularly effective for "vibe checks" and tonal consistency. Google describes this as an improvement in "nuance perception"—the ability to understand why a source said something, not just what they said.
With Edisource, you now have access to the industry's widest array of reasoning engines. We recommend testing Gemini 3 specifically on tasks where cultural context and tone are as critical as the raw facts—such as opinion editorials or localized reporting.
Gemini 3 also significantly outperforms on long form content including white papers, research reports and other analysis and insights content.
How to Access Gemini 3
Contact us today if you'd like to enable access to Gemini 3 in your Edisource account.
