Claude Fable 5 now available on Command Code. The Mythos class, state-of-the-art model shows exceptional performance in software engineering tasks, agentic workflows, and scientific research. Try it out now with /model

New in v0.33.2 Browse the web now from your terminal. Web Search → ranked results Web Fetch → reads full pages, not a snippet Training data has a cutoff. The web doesn't.

Command Code is always learning. Every accept, reject, and edit is a signal - auto-generates into project-level skills.

New in v0.32.3 ⋅ /configure-models Routes background requests to a model of your choice. Run compaction on MiniMax M2.5 Assign session titles with DeepSeek V4 Flash. Pick a model for each task, press r to reset. Defaults were a starting point. Now you have choices.

NVIDIA Nemotron 3 Ultra now available in Command Code! Strongest US open model yet! :four_leaf_clover: • 1M context • 5x faster inference • 550B MoE frontier-intelligence open model We've brought you the best pricing on the market for nemotron with ~500 TPS speed.

Stop making your workflow harder than it needs to be. `npm i -g command-code@​latest` You're welcome.

𝐘𝐨𝐮𝐫 𝐜𝐨𝐝𝐞. 𝐘𝐨𝐮𝐫 𝐭𝐚𝐬𝐭𝐞. The taste-1 model is the core of our taste architecture. To build your taste profile it: ▪ Learns from you ▪ Thinks like you ▪ Grows with you Learn more here: commandcode.ai/docs/taste

Command Code just crossed 10K paying customers! 🎉 It took 30 days to go from 0 to 10k customers. Fastest growing coding agent in history. Soon we'll be adding the same daily. Cooking something special here. Best harness for open models.

People keep asking if this is a typo. It’s not. commandcode.ai/pricing

Are you ready?!

Tomorrow Command Code will drop one of the best deals ever. Noon 12pm pacific. Guess and it's yours for free.

This is getting ridiculous! MiMo V2.5 is up to 99% off on Command Code. Let's get it!

Step 3.7 Flash is one of the fastest low-cost open models, now live in Command Code. Our $1 Go plan with 10x free usage credits, best way to try it. • 400 tokens/sec 🍃 • 256K context window • 3 reasoning levels Sharp instruction-following and strong /design taste.

Command Code x Claude Opus 4.8 Opus 4.8 is best for long-horizon agentic coding, large multi-file refactors, ambiguous specs that need pushback, complex multi-step agent runs, and high-autonomy work where reasoning quality matters most. • Same price as 4.7 • 1M context

Push & Pull Taste Share your coding taste with teammates, pull taste from the community.

MiMo-V2.5-Pro & MiMo-V2.5 are now ~99% off on Command Code. This is like 100x more usage! Input, output, and cache pricing are all lower. Works on every plan + extra top-ups. Pick the /model and go! Our $1 Go plan with $10 in it is perfect for this. Let's go!

Everyone said AI design was hopeless. We proved them wrong. Meet /design

Your AI coding agent finally has taste. Command Code's taste-1 model learns from every accept, reject, and edit you make, building a personal coding profile that gets smarter the more you use it. 10x faster coding. 2x faster reviews. 5x fewer bugs. commandcode.ai/docs/taste

Command Code now accepts UPI. Go plan billed in INR at ₹140/month. Card still works everywhere else. Pay in INR, code with 20+ models.

Saturday check-in: What are you working on right now? Show us a screenshot or drop a link

New in v0.26.21 --model now sets the model for any run. Works in the TUI. Works headless with -p. Fuzzy matching on every model ID. The right model for the job, picked at the prompt.

AI code works, but not like you. Devs waste time fixing the same patterns, rewriting tests/docs, waiting on PRs, and losing context between sessions. Command Code learns your style and applies it across projects. Make AI code like you → commandcode.ai

New in v0.26.20 Command Code supports Qwen 3.7 Max Frontier coding model with long-horizon agent execution. Try it out with /model

Introducing "/design": the skill that turns Command Code into your terminal-native design partner. One command. 16 modes. Each one knows what to look for, what to change, and what to leave alone. npm i -g command-code cmd /design help

A dollar for $40 of DeepSeek V4 Pro usage? Hard to say no to that.

/plan now takes a task inline. Reads your code before it writes any. Outputs the plan as a file. Edit it. Then run it. Execution is the easy part.

New in v0.26.11 Command Code supports Gemini 3.5 Flash. Pro-level coding, parallel agentic execution, 1M token context. Live in /model, same hour Google shipped it.

New in v0.26.5 /model now supports configurable reasoning effort. Set it from low through max. Auto for provider default. How hard your model thinks should be your call.

A CLI built for speed. And taste. ↳ Fast interactions ↳ Crisp outputs ↳ Beautiful terminal experience ↳ One-shot setup ↳ Start coding in seconds ↳ No configuration required Built for developers who care about craft. Get started: commandcode.ai

Take command. Of your code. Command Code helps you: • Ship features faster • Fix bugs • Write tests in your style • Refactor code • Learn your coding preferences over time Built for developers who want AI that adapts to how they work. Start here: commandcode.ai/docs

Imagine a coding agent… for $1/month. Stop imagining. CommandCode just dropped a $1 plan: ~15K requests $100 taste-1 usage ~$40 DeepSeek V4 Pro Analytics + Discord Yeah… $1. Start here ⟶ commandcode.ai/pricing

Your credits just got 4x stronger. DeepSeek V4 Pro is 75% off on Command Code until May 31: - Same budget, 4x the requests - Works on every plan + top-ups - Just pick the model, nothing else Don't leave credits on the table: commandcode.ai/pricing

Excited to announce our first video course Introduction to Command Code - 9 lessons. From scratch to shipping real apps. - Build software that matches your coding taste. - Live on our website and YouTube. Start for free ↳commandcode.ai/courses

Today we're shipping ten free design system taste packages. Inspired from industry best examples like: Figma, Vercel, Apple, Claude, Spotify, Uber Pull one of these and build UIs that look yours. Try now!! ↳ npx taste pull Read thread to download

Introducing Hooks in Command Code. Gives you deterministic control over agent behavior without relying on the model to follow instructions. - Block dangerous commands - Auto-format on edit - Run custom validations Try now! |_ npm i -g command-code

AI Terminology #28: Beam Search ↳ A decoding algorithm that explores multiple possible output sequences in parallel and keeps the most likely ones at each step to improve overall results.

People think coding agents only follow instructions. “Agents can’t think ahead.” Coding agents can plan next steps, anticipate issues, and iterate toward goals with minimal input. How have coding agents surprised you in your workflow?

AI use case: Environment setup AI can set up dev environments, install dependencies, and configure projects automatically. It gets you from zero to coding in minutes.

Announcing our first-ever Ship Week: April 20-24 We've been heads-down building, and can't wait to show what's coming. Five days. New launches every day. Stay tuned. P.S. Want a first look? We're inviting beta testers to get early access and share feedback.

People think coding agents reduce developer ownership. “Using agents means you’re less involved.” Coding agents increase leverage by handling execution while you stay focused on direction, quality, and product thinking. How have coding agents improved the way you lead development?

AI use case: Tool orchestration AI can coordinate your editor, terminal, tests, and documentation in one flow. It turns scattered coding tasks into a single, streamlined workflow.

AI Terminology #26: Temperature ↳ A decoding parameter that controls randomness in model outputs. Lower values make responses more deterministic, while higher values increase creativity and variation. x.com/CommandCode...

AI coding agents can now optimize code performance automatically. ↳ Profile bottlenecks from runtime traces ↳ Refactor hot paths for lower latency ↳ Re-test to verify speed gains Performance tuning is shifting from manual profiling to autonomous optimization.

People think coding agents are only for writing code. “They’re just advanced autocomplete.” Coding agents help with planning, debugging, testing, refactoring, and even coordinating full feature workflows. How are coding agents changing your workflow beyond coding?

AI use case: Context awareness AI can understand your codebase, dependencies, and current task before taking action. It helps every suggestion feel precise, relevant, and immediately useful.

AI Terminology #25: Softmax ↳ A function that converts raw model outputs (logits) into probabilities that sum to 1, helping the model choose the most likely next token.

AI coding agents can now write and validate tests automatically. ↳ Generate unit and integration tests ↳ Identify edge cases from code paths ↳ Ensure changes don’t break behavior Test coverage is becoming a default, not an afterthought. Follow for more insights on AI and coding agents.

People think coding agents need constant supervision. “You have to guide every step.” Modern agents can plan, take actions, and self-correct across tasks, letting you focus on higher-level decisions. How much do you let agents run on their own?

AI use case: Task planning AI can break complex coding tasks into clear, executable steps. It helps you move faster by turning ambiguity into structured action.

AI Terminology #22: Logits ↳ The raw, unnormalized output values from a model before probabilities are applied, typically passed through a softmax function.