AutoKaam Playbook
Gemini, Google's Surprisingly Strong Family
Gemini 3 Pro is the multimodal leader; Gemini 3 Flash is the cheap-and-fast bet.
Last reviewed:
The operator take
Gemini in 2026 is genuinely good, and that surprised me. The Gemini 1 era through 2023 was forgettable; Gemini 2 in 2024 was competitive on benchmarks but underwhelming in practice; Gemini 3 in 2025-26 is the first generation I have actually used in production.
Where Gemini 3 wins: multimodal grounded queries. Searching the web with Gemini's grounding feature returns cited, recent, accurate results in a way ChatGPT and Claude do not match. For research and fact-finding tasks, Gemini's grounded mode is genuinely useful, and I use it as a second opinion source on the empire stack.
The Pro tier (Rs 1,950 a month direct, plus the included Google AI Pro perks) is a real product. Free Drive storage, Gemini 3 Pro access in Gmail and Docs, and the unlimited Veo video tier (which is the cheapest path to AI-generated video at any quality I trust). For a knowledge worker living in Google Workspace, this bundles up to genuine value.
Where I do not use Gemini: production code work. Sonnet 4.6 still beats Gemini 3 Pro on the empire's coding benchmarks, and the gap is large enough that switching costs lose. The Gemini coding tools (Gemini Code Assist, the IDE integration) are competitive but not industry-leading.
The Indian-operator angle is the YouTube Premium bundling. The Google AI Pro plan in India often comes bundled with YT Premium, which is a genuinely useful tax-shelter for empire work that already needs Premium for ad-free YouTube. Worth checking your account for the bundle math.
The risk I hedge against: Gemini's API pricing has moved frequently through 2025-26. The Flash tier started at USD 0.075 per 1M tokens, climbed to USD 0.15, and has bounced. Building production infrastructure on Gemini Flash exposes you to per-month pricing whiplash that Anthropic and OpenAI have avoided.
Use Gemini for grounded research. Use it for video. Skip it for production code unless your stack is already Google-shaped.
Why it matters in 2026
Gemini 3 Pro is the strongest multimodal grounded model in 2026. For Workspace-heavy users, the Pro plan bundling delivers real value. Veo video remains the most-quoted AI video product in the consumer market.
Cost in INR
Free tier with Gemini 3 Flash. Google AI Pro: Rs 1,950/mo (often bundled with YT Premium in India). Ultra tier: Rs 19,500/mo (Veo unlimited, longer context).
Use when
- +Grounded web search with citations
- +Multimodal queries (image, video, document analysis)
- +Workspace integration for Gmail / Docs / Sheets users
- +AI video generation via Veo
Skip when
- xProduction coding workloads (Claude wins)
- xWhen pricing stability matters more than capability ceiling
- xLong-context reasoning where Opus 1M is the only viable choice
Alternatives I would consider
Read next
Adjacent in the playbook
Claude.ai Pro: ~Rs 1,700/mo (USD 20). Claude Max (5h sessions): ~Rs 8,500/mo (USD 100). API Sonnet 4.6: USD 3 input / USD 15 output per 1M tokens. API Opus 4.7: USD 15 input / USD 75 output per 1M tokens.
Claude, Anthropic's Sonnet and Opus Families
Free tier (limited GPT-5.4-mini). ChatGPT Go: Rs 399/mo (India-specific). Plus: Rs 1,700/mo. Pro: Rs 16,500/mo.
ChatGPT, the Default Most Indians Started With
Free tier available with Claude Pro. Claude Max (recommended for empire-scale work): ~Rs 8,500/mo. Pay-as-you-go API also supported.
Claude Code, the CLI Agent I Run All Day
Free tier with limits. Pro: Rs 1,700/mo (USD 20). Business: Rs 3,400/seat/mo (USD 40).