What AI coding assistants actually do

AI coding assistants do three jobs: autocomplete the line you're typing, answer questions about a codebase, and — increasingly — act as agents that edit files and run commands themselves. The productivity evidence is real: in the controlled GitHub Copilot study, developers completed a task 55.8% faster with the assistant than without [1]. The same study's nuance matters more than the headline: gains concentrate on routine implementation work, and review discipline is what keeps agent-made edits safe.

GitHub Copilot, decoded — including the "Copilot laptop" trap

Copilot's timeline: technical preview (June 2021), general availability at $10/month (June 2022), a free tier (December 2024), Pro+ at $39 (May 2025), a Max plan (May 12, 2026), and the switch to AI Credits usage-based billing on June 1, 2026 — fixed premium-request allotments were replaced by metered credits, and overages are billed [2][3]. Current pricing (September 2026): Free $0; Pro $10; Pro+ $39; Max $100; Business $19/user; Enterprise $39/user [4]. Microsoft reported Copilot passed 20 million users in August 2025 [5]. And the naming trap: searches for "review copilot laptop" or "lightest copilot laptop" are about Copilot+ PCs — Microsoft's branding for Windows laptops with NPU AI hardware — which have nothing to do with the GitHub Copilot coding tool. If a review mentions battery life and Snapdragon chips, it's a laptop; if it mentions code completion, it's the coding assistant.

Using Copilot in the terminal

"How to use copilot cli" and "how to use gh copilot cli" both point to GitHub's CLI integration: authenticate with gh auth login, install the Copilot CLI extension, and you get command suggestions and explanations in your shell — the "zsh ai copilot" workflow. The free tier includes limited CLI usage; heavier terminal use is what Pro+ is for.

The alternatives: Cursor, Claude Code, Windsurf, Amazon Q

Run AI locally: Ollama, LM Studio and the open models

"How to host ai locally" starts with two free runners. Ollama (first released July 2023, MIT-licensed) is the terminal-native runner with a huge model library — it hit nearly 9 million users and raised $65M in July 2026 [19][20]. LM Studio is the graphical alternative (built on llama.cpp/MLX): Free for local models, with paid tiers Bionic+ $20 and Pro $100 for US-hosted cloud inference with zero data retention [21]. Then pick an open-weights model:

The privacy payoff is the point: everything runs offline, nothing leaves your machine. The cost is capability — a local 8B model will not match ChatGPT or Claude on hard problems.

VRAM & model sizes: the honest numbers

As an editorial rule of thumb for "best lightweight coding llm" shopping: at 4-bit quantization you need roughly 0.5–1GB of VRAM per billion parameters. That means 8GB VRAM → 7–8B models, 16GB → ~14B, and 24GB+ → 32B comfortably. Llama 4.5 Scout's 70B-active MoE and Qwen3.8's 125B main model are workstation-class. When a laptop won't do, use a quantized model one size down rather than expecting a large model to run acceptably on undersized hardware.

What "temperature" actually means

"Best temperature llm" searches usually mean the model-sampling setting, not a model tier: temperature scales randomness in output, typically 0–2. Low values (0–0.3) give consistent, deterministic answers — right for coding, math and factual Q&A; higher values (0.7–1.0+) give varied, creative phrasing. It does not make a model "smarter" — a weak model at low temperature is just consistently weak. Our default recommendation: 0.2–0.4 for code, 0.7–0.9 for drafts.

Comparison table (verified 2026)

ToolFree tierEntry paidBest forWatch out
GitHub CopilotYesPro $10/moIDE integration, priceAI Credits metering since Jun 2026
CursorHobbyPro $20/moDedicated AI editorCredit caps push power users to Ultra $200
Claude CodeWith Claude FreeClaude Pro $20/moAutonomous agentsReview its edits; usage limits
WindsurfYesPro $15/moAgentic editor valueLimited free Cascade quota
Amazon Q DeveloperYes, 50 req/moPro $19/user/moAWS environmentsAWS-centric SSO/management
OllamaYes (MIT)—Terminal model runnerBound by your hardware
LM StudioYesBionic+ $20/moGUI local + cloud tiersFree tier web-search limits

Prices verified from vendor pages and reputable trackers, 2026. Sources: [4][6][10][11][15][17][21].

Frequently asked questions

GitHub Copilot vs Cursor vs Claude Code — which is best?+
Copilot wins on editor integration and price (free to $10/month); Cursor is the strongest dedicated AI editor (Pro $20); Claude Code is the most capable autonomous agent for advanced users on Claude Pro/Max [4][6][10].
Is GitHub Copilot free?+
Yes. Copilot has had a Free tier since December 2024; paid tiers run Pro $10, Pro+ $39, Max $100, Business $19/user, Enterprise $39/user. Since June 1, 2026, usage above plan allowances is billed in AI Credits [2][3][4].
How do I run AI locally on my PC?+
Install Ollama or LM Studio, then pull an open-weights model (Llama, Gemma, Qwen or Phi) sized for your GPU. 8GB VRAM suits 7–8B models; 16GB suits ~14B; 24GB+ suits 32B [19][21].
How much VRAM do I need to run Llama locally?+
Roughly 0.5–1GB per billion parameters at 4-bit quantization: 8GB for 7–8B models, 16GB for ~14B, 24GB+ for 32B. Llama 4.5 Scout's 70B-active MoE needs GPU-class hardware [23].
Ollama vs LM Studio — which is better?+
Both are free for local use. Ollama is the terminal-friendly runner with a huge model library; LM Studio has a GUI, a built-in model browser and optional paid cloud tiers. Choose by interface preference [19][21].
Is a "Copilot laptop" the same as GitHub Copilot?+
No. "Copilot+ PCs" are Microsoft's branding for Windows laptops with NPU AI hardware — unrelated to the GitHub Copilot coding assistant. The shared name causes constant confusion in reviews.
What does LLM "temperature" mean?+
Temperature controls randomness in model output. Low values (0–0.3) give consistent, factual answers — best for coding; higher values (0.7–1.0+) give varied, creative output. It does not make the model smarter.

Sources

  1. Peng et al., "The Impact of AI on Developer Productivity" — arxiv.org/abs/2302.06590
  2. GitHub Blog, "GitHub Copilot is moving to usage-based billing" — github.blog
  3. GitHub Blog, "Copilot individual plans: flex allotments and a new Max plan" — github.blog
  4. CloudZero, "GitHub Copilot cost" — cloudzero.com
  5. App Developer Magazine, "GitHub Copilot surpasses 20M users" — appdevelopermagazine.com
  6. Cursor, pricing — cursor.com
  7. KrAsia, "Cursor's $500M annualized revenue", Jun 2025 — kr-asia.com
  8. The Next Web, "Cursor (Anysphere): $2B funding, $50B valuation" — thenextweb.com
  9. TechCrunch, "Anthropic brings Claude Code to the web", Oct 2025 — techcrunch.com
  10. Anthropic support, "What is the Max plan" — support.claude.com
  11. Claude pricing — claude.com
  12. Le Monde, "Anthropic valued at $380 billion after a new funding round", Feb 2026 — lemonde.fr
  13. Anthropic docs, Claude Code overview — docs.claude.com
  14. The New York Times, "Cognition AI acquires Windsurf", Jul 2025 — nytimes.com
  15. DEV, "Windsurf AI pricing 2026" — dev.to
  16. AWS, "Amazon Q Developer generally available" — aws.amazon.com
  17. AWS, Amazon Q Developer pricing — aws.amazon.com
  18. Superblocks, "Amazon Q Developer pricing" — superblocks.com
  19. Wikipedia, Ollama — en.wikipedia.org
  20. TechCrunch, "Ollama raises $65M, grows to nearly 9M users", Jul 2026 — techcrunch.com
  21. LM Studio, pricing — lmstudio.ai
  22. Meta AI, "Llama 4: multimodal intelligence" — ai.meta.com
  23. MintedBrain, "Llama 4.5 Scout" — mintedbrain.com
  24. Google AI for Developers, Gemma releases — ai.google.dev
  25. Android Developers Blog, "Gemma 4: new standard for local agentic intelligence", Apr 2026 — android-developers.googleblog.com
  26. Geo.tv, "Google launches Gemma 4 open models: 400M downloads" — geo.tv
  27. Hugging Face, microsoft/phi-4 — huggingface.co
  28. Abdin et al., "Phi-4 Technical Report" — arxiv.org/abs/2412.08905
  29. Alibaba Cloud, "Alibaba introduces Qwen3" — alibabacloud.com
  30. Alibaba Cloud blog, "Qwen3.8-Flash-Next" — alibabacloud.com