Ampcode Release Notes

Opus 4.7 now powers Amp's smart mode.

In our internal evals, Opus 4.7 scored ~72%, up from Opus 4.6's ~65% - the first model since GPT 5.4 to clear 70%.

It takes some getting used to

Compared to Opus 4.7, Opus 4.6 was forgiving.

You could give it a vague task and it would often infer the missing pieces, make a plan, and start working. Sometimes that was useful. But it also could lead to the model confidently solving a nearby problem instead of the one you actually had. Or rushing to the first, but not the best, solution.

Opus 4.7 is less like that.

It follows prompts more closely. It fills in fewer gaps. It researches more. It is less likely to silently generalize from "fix this case" to "fix every related case." If the task is underspecified, you are more likely to get a narrow answer, a pause, or a request for the missing constraint.

At first, that can feel worse. But then you realize that a good prompt can make it go further.

Opus 4.7 is better at harder coding work, especially tasks that span multiple files, tools, and verification steps. It is better at keeping the shape of a change in its head and carrying it through the codebase. It's better at refactoring too. Its explanations are more thorough.

Fewer Built-in Tools

We removed grep and glob from smart.

Opus 4.7 is good enough at using the shell directly. When it needs to search, it can run rg or use the codebase search agent. …

When you use Amp Free, you won't see ads anymore.

We know many of you loved the ads, advertisers were happy, and we quickly grew ad sales to a $10M+ USD annual run rate, but the world has changed since we introduced ads in October 2025.

The launches of Gemini 3 Pro, Opus 4.5, and GPT-5.2 Codex starting in late November 2025 changed the world, and in this new world, ads don't make sense. Ads just don't pay for enough frontier tokens to make a difference, and token consumption is only going up from here.

Along with this, OpenAI now offers subscription plans with even more aggressive discounts than what Anthropic offered before. You can pay OpenAI $20/month to get (seemingly) $1000+/month in tokens. This gives you a lot more "free" usage than Amp's ad-supported free tier and is a better choice if cost is your top concern.

What about the $10 daily free usage of Amp? For now, most who have it will keep getting it, without ads. We'll be pausing it for some less-active users. As we ship updates to Amp, you can expect the free daily grant to be more available and more generous for people using Amp in the recommended ways, and less so for people using older Amp versions and workflows. We'll let you know before we make that change. Think of it as a bonus for staying on the frontier with us.

Update (2026-05-23): We’re pausing or reducing further Amp Free’s daily allowance for more users and restricting use of free credits by older client releases.

GPT-5.4 now powers Amp's deep mode.

It's the best model in the world right now.

It's faster than GPT-5.3-Codex and still as great at coding.

But, out of the box, GPT-5.4 was too chatty. That's not what we want for deep; it's not a pair programmer, it's supposed to go off and solve the problem.

So we tuned GPT-5.4 to behave like GPT-5.3-Codex.

Once we had that, we started to use it exclusively; even for interactive tasks. We run it at very high reasoning (deep^3) and still prefer it when we need fast interaction and fast reaction. It takes steering better than GPT-5.3-Codex.

To use it: open the command palette with Ctrl-O and run the command mode: use deep in the Amp CLI, or select deep mode in the Amp editor extension's prompt field. By default it uses deep^2, you can switch to deep^3 by hitting Opt-D.

Habemus oraculum! We have a new oracle in Amp and it's GPT-5.4.

It's a great model. In our internal evals response quality went from 60.8% (GPT-5.2) to 68.2% (GPT-5.4). Mean latency is down from ~6.7min to ~4.9min.

In Amp's smart mode GPT-5.4 works really well with Opus 4.6, which is smart mode's current main model. They complement each other with the oracle bringing sage advice on architecture, code reviews, and tricky bugs to the context window, just as we're used to from previous incantations.

On top of that, we also decided to add the oracle subagent to deep mode. Now you might wonder, since deep mode currently uses GPT-5.3-Codex as the main model, why add another GPT model in the same mode? Does that even make sense?

We think it does. GPT-5.3-Codex is fantastic at coding (as Codex models tend to be), which is exactly why it is the main model in deep, but the oracle is plain GPT-5.4, a non-Codex model. Less a code specialist, more an all-rounder.

That gives us two models from the same family, but trained for different goals, with different system prompts, in the same mode — two distinct voices in the same conversation.

We're still learning what GPT-5.4 can do in practice. There are very likely hidden smarts and treasures we haven't found yet. Let us know once you do.

We're closing admission to Amp Free, for now.

If you already have Amp Free, you’re still in, for now. You'll keep getting $10/day in free usage, supported by ads. If not, you'll need to keep paying for all of your Amp usage.

Why? Amp is growing very fast, but we need to slow down the growth because we're very busy building the next version of Amp. We want it to feel like the future of how you build software with agents.

We'll be sprinting on this frontier with our early-adopter customers and community. To do that well, we need to spend less time on new-user support and keeping usage fair (fraud and abuse prevention), and more time on building the next Amp.

When the next version of Amp is ready, you'll know, and we'll reopen Amp Free.

Code review has traditionally been tied to an interface where a human reads diffs. Now, we've fully decoupled the review agent from any UI, making it a composable and extensible subroutine that can be invoked from many different places where it is useful:

• You can run amp review in the CLI to run the review agent directly
• You can request a review in any thread using natural language like "review the outstanding changes" or "review changes since diverging from main"

This composability also means you can more easily close the loop by asking the main agent to automatically fix the issues found or by piping review comments into another command.

Invoking the review agent directly using amp review:

Requesting a review from within a thread:

Review tool

Customizing Review with Checks

You can also define Checks within your codebase. Checks are user-defined invariants or review criteria scoped to specific parts of your codebase. They are defined in .agents/checks/ directories.

Here's an example performance check, which you could save to .agents/checks/perf.md:

---
name: performance
description: Flags common performance anti-patterns
---

Look for these patterns:

- Nested loops over the same collection (O(n²) → O(n) with a Set/Map)
- Repeated `array.includes()` in a loop
- Sorting inside a loop
- String concatenation in a loop (use array + join) …

We're removing custom commands in favor of skills. They were two ways of doing the same thing, except that only you could invoke custom commands, and only the agent could invoke skills. Now that you can invoke skills directly, there's no reason to keep custom commands around anymore.

Migrating Custom Commands to Skills

It's easy and takes just a few minutes.

Let Amp Do It For You

If you have custom commands in .agents/commands/ or ~/.config/amp/commands/, ask Amp to migrate them for you with the following prompt:

Migrate my custom commands from .agents/commands/ to skills in .agents/skills.

For each command:
1. Create a new skill directory in .agents/skills/ named after the command
2. If it's a markdown file, rename it to SKILL.md and add frontmatter with name and description
3. If it's an executable, create a SKILL.md that describes when and how to run the script, and move the executable to a scripts/ subdirectory
4. Delete the original command file

Also do this for global custom commands in ~/.config/amp/commands (which should be migrated to skills ~/.config/agents/skills).

Or Do It Manually

Markdown commands: Move your .md file to a skill directory:

# Before
.agents/commands/code-review.md

# After
.agents/skills/code-review/SKILL.md

Add frontmatter:

---
name: code-review
description: Review code for quality and best practices
---

[Your existing content]
``` …

(Update on February 19, 2026: deep now uses GPT-5.3-Codex.)

Amp has a new agent mode: deep. It uses GPT-5.2-Codex and a selection of tools that are tuned to what this model is good at.

Where smart is eager, collaborative, and happy to edit right away, deep will silently read files and move through the codebase for five, ten, sometimes fifteen minutes before making changes - happy to be left alone.

GPT-5.2-Codex in deep is not an assistant that's constantly checking in with you. It wants to go off and solve problems on its own, not pair program. But that requires a clear definition of the problem up front, which you'll either need to provide in your first prompt, or ask deep to work out with you before you tell it to go and implement it.

Here are some prompts that worked really well in deep mode:

• Nicolay, changing how Skills register MCP tools: "A user has multiple skills, each skill has a mcp.json with the same server name and url, but with different includeTools list. So each skill uses a different subset of tools. Each mcp can only be registered once and last registration wins. But for each mcp within a skill we are only registering its subset of tools. So when the user has multiple skills, it is probably only having one set of tools being registered instead of all. How can we fix this? we could: [...]" …

We're removing Amp Tab. It is not part of the future that we see.

A year ago, most of our code was written by hand. In June, when we released Amp Tab, Amp was already writing the majority of our code. But now, Amp writes 90% of what we ship.

Amp Tab and other completion engines come from a world in which everyone believed humans will write most of the code and AI is sprinkled on top.

But that world is dying! Look around! Some of our users are saying they haven't opened their editor in days and yet still shipped code. The bottleneck is now how fast you can get the code out, not how fast you can write it, they say.

The era of tab completion is coming to an end. We're entering the post-agentic age, in which it's a given that agents write most of the code.

There's so much to figure out, so much to build, so much to explore! We have to choose what to focus on.

We're focusing on what's coming next, not what brought us here.

Amp Tab will continue to work until the end of January 2026. After that, we can recommend Cursor, Copilot, or Zed if you need inline completions.

Here's Quinn and Thorsten on the end of Amp Tab and how they feel about it (Quinn will miss it):

Amp can now generate and edit images, which is useful for design inspiration, tweaking mockups, and making visual assets:

• Exploring UI alternatives: "show me 3 different designs for this task dashboard"

• Editing an existing image: "update this dashboard with the changes shown in the attached image"

To use it, ask Amp to use the painter tool explicitly. The painter uses Gemini 3 Pro Image (a.k.a. Nano Banana Pro) under the hood.

Update (2025-05-06): Four months after this was posted, we removed handoff. So, if you want to fork a thread, the other recommendation here stands: make a new thread and mention the original thread by URL or ID.

We're ripping out the Fork command.

We added thread forking back in July 2025 (now ancient, primordial history) as a way to conveniently share context for branching experiments or side quests in Amp.

Today we have better ways of sharing context between threads: handoff and thread mentions, which treat threads as first-class stores of context.

Perhaps there is a great potential UX out there for fork, but we want Amp to be simple as well as powerful. We'd rather spend our time perfecting handoff and thread mentions than support fork.

Handoff

Handoff is great for extracting useful context from your thread for the next goal at hand. This means you can start a new thread with only the necessary context.

• Use thread: handoff from the command palette.
• Prompt your task and a new thread will be started with the necessary context already in the prompt.

Thread Mentions

Thread mentions let you pull information from other threads into your current thread. You can reference multiple threads, merging context from many sources. …

Agents Panel

The Amp editor extension now has a new panel to view and manage all active agent threads.

You can use the keyboard to navigate between threads:

• j/k or arrow keys to move between threads
• Space to expand a thread panel to show the last message or tool result
• Enter to open a thread
• e to archive or unarchive a thread
• Esc to toggle focus between the thread list and the input, which starts new threads

We recommend archiving old threads so the displayed threads represent your working set.

As coding agents improve and require less direct human oversight, more time will be spent by humans in managing and orchestrating work across multiple agent threads. We'll have more to share soon.

To get started, click the button on the left end of the navbar or use Cmd-Opt-I (macOS) or Ctrl-Alt-I (Windows/Linux).

Update: Amp Free no longer shows ads. See the latest update.

Every Amp user can now receive daily free credits to use the full Amp experience, including our frontier smart agent powered by Opus 4.5. No payment required, powered by ads—turn them off if you don't want free credits.

If you're a new user, just sign up for Amp, and download the CLI or editor extension.

If you're an existing user, go to user settings to enable ad-supported free credits.

What You Get

The free credit grant replenishes hourly, giving you a total of $10 worth of credits per day or roughly $300 of credits per month. When you've used up free credits, you can either wait for the hourly reset or purchase paid credits.

Amp's smart mode is currently driven by Opus 4.5 (with GPT-5 and Gemini-3-powered subagents like the oracle + librarian). You can also use rush mode, which provides faster inference at a lower cost per token, currently driven by Haiku 4.5.

Really?

Really! Like everything we do, ad-supported inference is an experiment. We can't promise we can do this forever, but we've already rolled it out to a sizable beta group thanks to our ad partners and it has been well-received. …

MCP servers often provide a lot of tools, many of which aren't used. That costs a lot of tokens, because these tool definitions have to be inserted into the context window whether they're used by the agent or not.

As an example: the chrome-devtools MCP currently provides 26 tools that together take up 17k tokens; that's 10% of Opus 4.5's context window and 26 tools isn't even a lot for many MCP servers.

To help with that, Amp now allows you to combine MCP server configurations with Agent Skills, allowing the agent to load an MCP server's tool definitions only when the skill is invoked.

How It Works

Create an mcp.json file in the skill definition, next to the SKILL.md file, containing the MCP servers and tools you want the agent to load along with the skill:

{
	"chrome-devtools": {
		"command": "npx",
		"args": ["-y", "chrome-devtools-mcp@latest"],
		"includeTools": [
			// Tool names or glob patterns
			"navigate_page",
			"take_screenshot",
			"new_page",
			"list_pages"
		]
	}
}

At the start of a thread, all the agent will see in the context window is the skill description. When (and if) it then invokes the skill, Amp will append the tool descriptions matching the includeTools list to the context window, making them available just in time. …

Amp has a new agent and it specializes in code review.

In the VS Code extension, you can use this agent by going to the review panel. Start by dragging the selection of changes you want to review. Sometimes, you want to review a single commit; other times you want to review all outstanding changes on your branch:

Amp will pre-scan the diff and recommend an order in which to review the files. It will also provide a summary of the changes in each file and the changeset overall:

This addresses a key difficulty in reviewing large changesets, as it's often difficult to know where to start.

Clicking on a file will open the full file diff, which is editable and has code navigation if your diff includes the current working changes:

Review Agent

The review agent lives in a separate panel below. It analyzes the changes and posts a list of actionable improvements, which can then be fed back into the main Amp agent to close the feedback loop:

…