$ diff gemini-3-6-flash grok-build
Gemini 3.6 Flash vs Grok Build
Google DeepMind and xAI, side by side. Every value below is the one recorded on the model’s own wiki page, with its citation — nothing is estimated.
Google DeepMindGemini 3.6 Flash
- context
- 1.0M
- weights
- closed
- $/M in
- $1.5
- $/M out
- $7.5
- context
- 256K
- weights
- open
- $/M in
- $1
- $/M out
- $2
What actually differs
- Context window
- Gemini 3.6 Flash takes 1.0M against 256K — 4.1× more room in a single request.
- Input price
- Grok Build at $1/M against $1.5/M — 1.5× cheaper to feed.
- Output price
- Grok Build at $2/M against $7.5/M — 3.8× cheaper to generate. Output dominates the bill on most agentic workloads, where the model writes far more than it reads.
- Weights
- Grok Build publishes weights (Apache 2.0 (CLI codebase, open-sourced 2026-07-16)); Gemini 3.6 Flash is API-only. That decides self-hosting, air-gapped deployment and fine-tuning before any capability question does.
- Recency
- Gemini 3.6 Flash shipped 68 days after Grok Build (2026-07-21 vs 2026-05-14).
Full spec
| Attribute | Gemini 3.6 Flash | Grok Build |
|---|---|---|
| Developer | Google DeepMind | xAI |
| Released | 2026-07-21 | 2026-05-14 |
| Context window | 1,048,576 tokens (1M) | 256K tokens |
| Pricing | $1.50/M input · $7.50/M output | $1/M input · $2/M output (xAI API public beta) |
| License | proprietary | Apache 2.0 (CLI codebase, open-sourced 2026-07-16) |
| Availability | Gemini API, Google AI Studio, Android Studio, Vertex AI, consumer Gemini app, Google Search, GitHub Copilot | Terminal CLI (SuperGrok Heavy), Kilo Code VS Code/JetBrains (SuperGrok or X Premium+), xAI API public beta; source at github.com/xai-org/grok-build |
Values come from Gemini 3.6 Flash and Grok Build, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing. How these pages are produced.