Reference Library
AutoKaam Playbook
30 tools, reviewed by an operator. INR pricing, the spec sheet I actually run, and the one thing each tool gets wrong.
Last reviewed across the whole library: 2026-05-06. 30 entries live.
Dev Tools
10 entriesClaude Pro: USD 17/mo on annual billing (USD 200 up front) or USD 20 month to month. Claude Max: USD 100/mo for 5x Pro usage, USD 200/mo for 20x (web prices). API per 1M tokens: Sonnet 5 USD 2 in / 10 out, Opus 5.5 USD 4 / 20, Fable 5.1 USD 10 / 50, Haiku 4.5 USD 1 / 5. Checked 28 Sep 2026.
Claude, Anthropic's Sonnet and Opus Families
The model I reach for when reasoning matters more than throughput.
Read playbookIncluded with Claude Pro (USD 17/mo on annual billing, USD 20 month to month) and Claude Max (USD 100/mo for 5x, USD 200/mo for 20x). Not included on the Free plan. Or pay per token through the API. Checked 28 Sep 2026.
Claude Code, the CLI Agent I Run All Day
Anthropic's terminal-first coding agent, and my daily driver for code work.
Read playbookFree tier with limits. Pro: Rs 1,920/mo (USD 20). Teams: Rs 3,840/seat/mo (USD 40).
Cursor, the IDE I Tried and the Empire's Soft Pass
Genuinely good IDE-pane AI; I just prefer the terminal flow for empire work.
Read playbookFree and open source (Apache-2.0). API costs at vendor rate. Ollama-backed runs at zero marginal cost.
Cline, the VSCode Agent That Stays Out of My Way
The free open-source alternative to Cursor's agent mode, with bring-your-own-key.
Read playbookFree and open source (Apache-2.0). API costs at vendor rate, and scale with how much you use it.
Aider, the Terminal Pair-Programmer With a Soul
git-aware command-line AI coding; the original good open-source agent and still excellent.
Read playbookFree tier available. Paid tiers step up from a lower-cost India-specific plan (Go) through Plus to a Pro tier for power users.
ChatGPT, the App Many Indians Started With
A common first AI app in India, and a defensible choice for general-purpose work.
Read playbookGPT-6 Astra: USD 10/50 per 1M tokens (input/output, standard short-context tier). GPT-6 Sol: USD 2/10. GPT-6 Luna: USD 0.10/0.50.
OpenAI API, the Developer Surface I Use Sparingly
GPT-6 Luna for cost-sensitive batch work; structured output is the genuine win.
Read playbookFree tier with Gemini 3 Flash (India: Rs 0). Google AI Plus: Rs 399/mo. Google AI Pro: Rs 1,950/mo (bundles YouTube Premium Lite in India). Google AI Ultra: Rs 6,500 to Rs 19,500/mo depending on usage tier (Veo 3.1 video, up to 1M token context).
Gemini, Google's Surprisingly Strong Family
Gemini 3 Pro is the multimodal leader; Gemini 3 Flash is the cheap-and-fast bet.
Read playbookGemini 3.1 Pro (Preview): USD 2.00 / 12.00 per 1M tokens (standard tier, prompts up to 200k tokens). Gemini 3.8 Flash: USD 0.75 / 3.75, rising to USD 1.50 / 7.50 from 1 Jan 2027. Free tier available via AI Studio for development. Checked 28 Sep 2026.
Gemini API, the Google Developer Surface
Cheap, fast, multimodal-strong; production-viable with caveats around pricing volatility.
Read playbookDeepSeek-V4.1-Flash: USD 0.30/1.20 per 1M tokens (peak), USD 0.15/0.60 (off-peak). DeepSeek-V4-Pro: USD 1.32/3.96 (peak), USD 0.66/1.98 (off-peak). Off-peak is exactly half of peak. Available via OpenRouter at a markup. Checked 28 Sep 2026.
DeepSeek, a Cheap Reasoning Tier I Trust
DeepSeek-V4.1-Flash and V4-Pro are the current models, priced per 1M tokens with a built-in off-peak discount.
Read playbookLocal LLMs
10 entriesFree, open source. Compute cost on consumer hardware is electricity, roughly Rs 4 to Rs 8 per active inference hour on a 65W desktop.
Ollama, the Local Model Runtime I Actually Trust
One binary, one model registry, zero cloud dependency. The default I reach for first.
Read playbookFree, open source. Compile-time cost on a M75q is under two minutes, on a Pi 4B about ten minutes.
llama.cpp, the Engine Under Most Local Inference
Compile once, run anything. Where I go when Ollama does not expose the knob I need.
Read playbookDesktop app: free for personal and internal business use (per LM Studio's own terms). Bionic cloud inference: Free tier, Bionic+ USD 20/mo, Pro USD 100/mo.
LM Studio, the GUI On-Ramp for People Who Hate Terminals
Polished desktop app for local models. I do not run it, but it converts non-developers fast.
Read playbookFree, open source (Apache-2.0). Compute cost via RunPod is about Rs 50 per hour for L4 (24GB VRAM, Community Cloud) and roughly Rs 150 to Rs 330 per hour for A100 or H100 (Community Cloud).
vLLM, Serving Throughput That Defends a GPU Bill
Production inference for teams. I only run it on RunPod, never local, and the math works.
Read playbookFree, open weights. Compute cost is local hardware electricity, effectively zero for personal use.
Gemma, the Open Family I Actually Reach For
Google's open-weight line. The 2B is my Pi default; the 9B is my desktop default; vision is the one I run daily.
Read playbookFree open weights. Cerebras: new accounts get a one-time USD 5 trial credit (card required, expires in 30 days); qwen-3.8-27b is limited to 5 requests per minute on that trial, then scales with paid usage.
Qwen, Where Cerebras Speed Plus Open Weights Actually Compose
Alibaba's family; Cerebras serves it fast, though the free access is a time-boxed trial, not a forever quota.
Read playbookA one-time empire credit grant ran from 28 Apr to 28 May 2026 (expired). Current direct-API rates (cache miss, updated 22 Sep 2026): mimo-v2.6-pro USD 0.435/0.87 per 1M tokens (in/out), mimo-v2.6-flash USD 0.14/0.28.
Xiaomi MiMo, a Cheap Grunt LLM for the Empire Stack
mimo-v2.6-pro and mimo-v2.6-flash, priced per 1M tokens now that the credit grant has run out.
Read playbookFree open weights (MIT). Compute cost on consumer hardware is unfavorable above 8B-class. Hosted API (off-peak) is roughly Rs 10 per 1M input tokens, Rs 60 per 1M output for V4.1-Flash.
DeepSeek Local, the Pricing Disruptor I Mostly Run Hosted
V4 weights are open. I downloaded them, learned the lesson, went back to the API.
Read playbookFree, open source. Compute cost on consumer hardware is electricity, effectively zero per hour of audio.
Whisper, the Local Transcription I Run on Every Voice Memo
OpenAI's open ASR model. Powers the empire field-note pattern. Zero rupees per hour of audio.
Read playbookFree, open source. CPU image-gen is real-time-uneconomic. RunPod L4 batch is roughly Rs 14 per 200 images including warm-up.
Diffusers, the Image-Gen Path I Use Sparingly
Hugging Face's library. Honest about what does not work without a real GPU.
Read playbookWorkflows
4 entriesFree self-hosted; Cloud tier from Rs 2,180/mo (EUR 20, billed annually; 2,500 executions/mo)
n8n, Workflow Automation Without the Cloud Tax
Self-hostable, fair-code, and far cheaper than Zapier once you cross 1,000 runs a month.
Read playbookFree self-hosted; Cloud free tier 50K units/mo; Pro from Rs 19,080/mo (USD 199)
Langfuse, the LLM Observability Stack I Actually Run
Self-hosted tracing for prompts, latency, cost, and eval results, with no per-trace pricing.
Read playbookFree open source; LangSmith (paid sister product) from Rs 3,740/mo (USD 39)
LangChain, the Framework I Have a Love-Hate Relationship With
Powerful when you need it, overkill when you don't, and the breaking-changes tax is real.
Read playbookFree open source; Enterprise tier is custom-priced (contact sales)
CrewAI, the Multi-Agent Framework I Tried Three Times
Clean conceptual model, weak production story, but the cleanest mental scaffold I have found.
Read playbookInfra
6 entriesFree open source; running cost depends on host (Rs 1,200/mo Oracle ARM via Coolify is empire baseline)
FastAPI, the Empire's Default Python Serving Layer
Async by default, OpenAPI for free, and the framework I reach for first when I need an HTTP endpoint.
Read playbookFree open source; running cost depends on host (Rs 1,200/mo Oracle ARM via Coolify supports 5+ empire apps)
PocketBase, the Empire Backend I Run Across Every Project
SQLite plus auth plus realtime plus admin UI in one binary; I have not regretted picking it once.
Read playbookFree open source; Oracle ARM free tier covers 2-core 12GB; Rs 1,200/mo for paid scale-out instances
Coolify on Oracle ARM, the Empire Hosting Stack
The Heroku-style PaaS I actually run, on the cheapest serious cloud bare-metal in 2026.
Read playbookFree at entry; Pro from Rs 1,920/mo (USD 20, billed annually; only needed for advanced WAF / analytics)
Cloudflare Pages, the Empire Static Hosting I Have Never Regretted
Free static hosting, fast global edge, and the autokaam.com deploy backbone.
Read playbookFree at 100K daily requests; Paid from Rs 480/mo (USD 5) for higher volume + larger CPU budgets
Cloudflare Workers, the Edge Runtime I Use for the Sharp Bits
100K free daily requests, no cold starts, and the empire's ads-txt + redirect layer.
Read playbookFree Tier: 500 Document Transactions a month (Adobe). Paid volume is priced on Adobe's developer site.
Adobe PDF Services, My PDF Backbone
Adobe's PDF engine as an API: extract, OCR, compress, protect and convert PDFs, with a free tier of 500 transactions a month.
Read playbook