AI Model Comparison
Side-by-side comparison of pricing, context window, benchmarks, capabilities, and license. Every figure comes from the developer's own pages, last checked on ; each model's page lists its sources. A figure we could not source shows as a dash.
Model A — Aleph Alpha
Kolibri is a mixture-of-experts reasoning model from Aleph Alpha for German and English, released on October 3, 2026 and built for public administration and industry, with deployment on customer-controlled infrastructure. It supports a reasoning mode and tool calling.
Model B — Google DeepMind
Gemini 4 Carbon is a Google model, available to paying developers in the Gemini API from November 10, 2026, which Google describes as its frontier model for long-horizon agentic work and its strongest in cybersecurity. Based on Gemini 4 Argon, it takes text, image, audio and video input, returns text, and has a March 2026 knowledge cutoff.
| Property | Kolibri | Gemini 4 Carbon |
|---|---|---|
| Released | Oct 3, 2026 | Nov 10, 2026 |
| Developer | Aleph Alpha | Google DeepMind |
| Type | reasoning | reasoning |
| License | open-weight | proprietary |
| Context window | 1,048,576 | — |
| Parameters | 78B total, 3.46B active per token (Mixture-of-Experts) | — |
| Developer status | — | — |
| Input $/M tokens | — | $2.00 |
| Output $/M tokens | — | $10.00 |
Bold green values indicate the winner on that dimension where comparison is meaningful (lower price wins; bigger context wins; more open license wins; newer date wins). Prices are the ones each developer lists today; a model it no longer prices has none.
Programmatic access available at /api/ai/compare?a=aleph-alpha-kolibri&b=gemini-4-carbon. CORS is open. Citation guidance: /about/editorial-process.