Use Lookout
Models and providers
Connect the providers you have a plan with or add a custom provider, then pick a model, its effort and its variant for each chat.
Lookout runs on plans you already have. A provider is the company you hold the account with (OpenAI for a ChatGPT plan, Anthropic for a Claude plan or a Claude API key); you connect it once with its own sign-in, and its models then appear in the model menu of every chat. OpenCode's free models are on from the start, so a chat can run before you connect anything.
Connect a provider
The provider moves to Connected and its models join the model menu. Click a provider's row to open its page: Accounts (each account and what its plan has left), Plan use (with two or more accounts, or extra usage) and Models. Providers in the page's title, or Escape, goes back to the list. Anthropic's page lists its Claude plans and its API key together, then each one's plan use and models under its own name.
These sign-ins happen in the app. Your Lookout account carries them to your other machines, so a machine with only the CLI gets them once it is on your account: Providers on this machine.
Sign out on an account signs it out on all your devices and on the web dashboard, whether it still works or its sign-in expired. An expired sign-in your account holds reads Sign-in expired on every device's Providers page, even on one that never had it, with Sign in again and Sign out, and counts there as it does on the dashboard. It stays listed until you do one or the other.
On a machine with only the CLI
With no Lookout account, or before the machine is on it, the CLI still runs:
With a code
SpaceXSI (Grok), Moonshot AI (Kimi) and Meta (Muse Code). Lookout opens the provider's page and the row shows Your code. Enter the code on that page and approve. Copy code and open the page copies the code and opens the page again. The code expires: the row says when.
In the browser
Google (Antigravity), OpenAI (ChatGPT), Anthropic (Claude) and Command Code. For your first account Lookout opens the provider's sign-in page, and the row says Finish in your browser. until you do. Open it again opens the page again. When the provider already has an account, the page doesn't open by itself: the step offers Open sign-in and Copy link for a private window side by side, because the page signs in whoever your browser is already signed in as. When Lookout knows which account the step is for, it says so, for example Approve as [email protected] on claude.ai.
The page signs in whoever your browser is already signed in as. For another account, click Copy link for a private window (or Copy the link on a first sign-in) and open it in a private window or another browser profile on the same Mac (Add another account). If the page signs in an account you already added, nothing is added and the page says so: choose Add account again and use Copy link for a private window.
From another device
OpenAI (ChatGPT), Google (Antigravity) and Anthropic (Claude plan) can be finished in a browser on any device, such as your phone or a friend's computer. Click Use another device under the step, then Copy the link, and open it in that browser and approve:
Lookout tries what you paste at once and adds the account. The code or address works for this sign-in only, once: if it has expired or was already used, open the link again. A sign-in with a code works from any device as it is. Command Code's sign-in finishes in a browser on the machine running Lookout only.
When the Command Code CLI is signed in on this device, the row offers that sign-in first: click Use this account, or Sign in with another account.
Every Claude account in Lookout is Lookout's own sign-in: Lookout renews it itself, it syncs to your other devices,
and it keeps working whatever Claude Code is signed in to, so you can switch Claude Code between your subscriptions
as often as you like. When Claude Code is signed in to an account you haven't added, Sign in with Claude offers it: Add
[email protected] starts the sign-in for that account, and the step says Approve as [email protected] on
claude.ai. If your browser is signed in to claude.ai as another account, use Copy link for a private window,
sign in to claude.ai there as that account, and approve. Use one private window per account, and close it before the
next. If you keep one Claude Code login per account (a CLAUDE_CONFIG_DIR each, such as
CLAUDE_CONFIG_DIR=~/.claude-work claude), Lookout offers each of their accounts in turn. Signing an account out in
Lookout doesn't affect Claude Code.
Lookout never reads Claude Code's sign-in, only which account it is signed in to. A Claude account an older version borrowed from Claude Code no longer works: add it again with Sign in with Claude. A Claude sign-in lasts about a month; three days before it ends, its row says Sign-in ends with the date and offers Sign in again, and keeps working until then.
With an API key
Anthropic, DeepSeek and Z.ai. Click Add API key and paste your key. Get a key opens the provider's page that issues keys. Lookout checks the key with the provider, and a key the provider refuses is not kept.
Anthropic is one provider with two kinds of access: a Claude Pro, Max or Ultra plan (Sign in with Claude, bought
on claude.ai) and an API key from the Claude Console (Add API key, billed per token). Either works alone, or both
together, and the Anthropic row lists every account of both, each naming its kind (Claude Ultra, API key). Its
Add account offers Sign in with Claude first and Add API key second; the key goes in a dialog whose Get a
key opens the Claude Console's API keys page. The key's models have their own IDs (anthropic/ and the model's
name), so a chat says which bill it runs on, and where the model menu lists a model on both, the API key's row ends with API key and each row's
tooltip says On your Claude plan or On your API key. The provider reports no balance for a normal key, so the key's line shows budget
left: the budget you set minus what Lookout counted as spent on this device, at Anthropic's list prices (Read a
plan's meters). That is not Anthropic's remaining credit. To bring in a key you keep in the macOS
Keychain without pasting it:
lookout providers import-key anthropic --keychain SERVICE [--keychain-account ACCOUNT] [--models claude-haiku-5-5] [--budget 100]The key is read from the Keychain item, checked with Anthropic and kept like a pasted one, and it is never printed.
--models limits the account to those models (the others stay listed as not enabled for this key), and
--budget stops it once its spend reaches that many US dollars.
With another app's sign-in
Cognition's Devin signs in through its own command-line app. When that app is signed in on this device, Sign in with Devin uses its sign-in. When it isn't, the row shows the command to run first: run it in a terminal and sign in, and Lookout picks the sign-in up by itself.
| Provider | App | Command |
|---|---|---|
| Cognition | The Devin CLI | devin auth login |
A free tier
OpenCode needs no sign-in, and its free models are on from the start. To take them out of the model menu, open the OpenCode row and choose Remove from its menu. Add brings them back. This choice stays on the device you make it on. OpenCode's free models run in chats only: subagents and workflow agents are not offered them.
The list follows OpenCode's own: Lookout reads the model catalog the OpenCode app reads (models.dev), so a model OpenCode makes free joins the model menu, with its context, output limit and effort levels as OpenCode lists them, within the hour and without a restart, and one it retires leaves it. A chat set to a retired model keeps its history, and its next message says the model left OpenCode's free models and names those free now; Continue on under it moves the chat. Names show as OpenCode writes them. When the OpenCode app is signed in with an OpenCode Go key, Go's free models show too, marked (Go); they run on that key.
When OpenCode refuses Lookout's client on a model, the chat that hit it says so and that model reads unavailable on that device until a request to it succeeds again. When it refuses every free model, the row in Settings › Providers reads Refused Lookout's client since the day it began, with what OpenCode said on its page.
Providers
| Provider | What you need | Sign-in | Accounts | Web search |
|---|---|---|---|---|
| Anthropic | A Claude Pro, Max or Ultra plan, a Claude API key billed per token, or both | In the browser, API key | Up to 15 of each | Yes |
| Cognition | A Devin seat | The Devin CLI | 1 | Yes |
| Command Code | A Command Code plan | In the browser | Up to 15 | No |
| DeepSeek | A prepaid DeepSeek API balance | API key | Up to 15 | No |
| A Google Antigravity plan | In the browser | Up to 15 | No | |
| Meta | A Muse Code subscription | Code | Up to 15 | Yes |
| Moonshot AI | A Kimi Code membership | Code | Up to 15 | No |
| OpenAI | A ChatGPT plan | In the browser | Up to 15 | Yes |
| OpenCode | Nothing: free models from OpenCode Zen | None | None | No |
| SpaceXSI | A paid X Subscription plan, or a SuperGrok plan | Code | Up to 15 | Yes |
| Z.ai | A Z.ai Coding Plan | API key | Up to 15 | Yes |
With web search, the agent can search the web while it works. On every model it can still read a web page whose address it has. Several accounts on one provider: Accounts and usage.
Custom providers
A model host that isn't in the table, such as a hosted model with its own API or a server on your Mac, is a custom provider. In Settings › Providers, Add provider under Custom providers asks for a name, the base URL, the API format it speaks (Chat Completions, Responses or Anthropic Messages) and its key, which stays hidden unless you show it. Lookout checks the key with the host's model list before it keeps anything, then offers that list: check the models you want. A model the host doesn't list goes in with Add model.
Each model has its ID, the name the model menu shows, its context window and the most a reply may use. Smart configuration fills them from the host's list, or from Lookout's own catalogs when the model is one they know. Advanced settings hold the rest: image input, the host's own web search, the reasoning levels it takes, and the JSON that carries the chosen level in each request. Test sends one small request and says whether it went through.
A custom provider's models appear in the model menu under its name. Its usage counts tokens only: Lookout knows no
plan or price for it. The provider and its key sync to your other devices with your account, sealed in its vault: the
web dashboard counts its tokens under "custom" and never sees its name, address or models. On a machine with only
the CLI, lookout providers custom add does the same (CLI).
Models
A new install lists these models. Once you sign in, the list follows your account: a plan can add a model or a
context window, a model your plan or key can't run says why under its name, such as Not on your plan or Not
enabled for this key, and one the provider is retiring shows
Retires and the date. Settings › Providers lists what your account has. The ID is what lookout exec -m
takes.
| Provider | Model | Variant | Context | Effort | ID |
|---|---|---|---|---|---|
| Anthropic, API key | Fable 5.2 | 1M | Low to Max | anthropic/claude-fable-5-2 | |
| Anthropic, API key | Haiku 5.5 | 1M | Low to Max | anthropic/claude-haiku-5-5 | |
| Anthropic, API key | Mythos 5.2 | 1M | Low to Max | anthropic/claude-mythos-5-2 | |
| Anthropic, API key | Opus 5.5 | 1M | Low to Max | anthropic/claude-opus-5-5 | |
| Anthropic, API key | Sonnet 5.5 | 1M | Low to Max | anthropic/claude-sonnet-5-5 | |
| Anthropic, Claude plan | Fable 5.2 | 1M | Low to Max | claude-fable-5-2 | |
| Anthropic, Claude plan | Haiku 5.5 | 1M | Low to Max | claude-haiku-5-5 | |
| Anthropic, Claude plan | Mythos 5.2 | 1M | Low to Max | claude-mythos-5-2 | |
| Anthropic, Claude plan | Opus 5.5 | 1M | Low to Max | claude-opus-5-5 | |
| Anthropic, Claude plan | Sonnet 5.5 | 1M | Low to Max | claude-sonnet-5-5 | |
| Cognition | Gemini 4 Argon | 1M | Low to High | gemini-4-argon | |
| Cognition | SWE-2 | 262K | Medium to Max | swe-2 | |
| Command Code | DeepSeek V4.1 Flash | 1M | None to Max | deepseek-v4.1-flash | |
| Command Code | MiMo V2.6 Pro | 1M | Not adjustable | mimo-v2.6-pro | |
| DeepSeek | DeepSeek V4.1 Flash | 1M | None to Max | deepseek-flash | |
| Gemini 3.8 Flash | 1M | Low to High | gemini-3.8-flash | ||
| Gemini 4 Argon | 1M | As your plan lists | antigravity/gemini-4-argon | ||
| Gemini 4 Carbon | 1M | As your plan lists | antigravity/gemini-4-carbon | ||
| Gemini 4 Hydrogen | 1M | As your plan lists | antigravity/gemini-4-hydrogen | ||
| Meta | Spark 1.3 | 1M | Minimal to Extra high | muse-spark-1.3 | |
| Meta | Spark 1.3 | Contributor | 1M | Minimal to Max | muse-spark-1.3-contributor |
| Meta | Wonder 1.0 | 1M | Minimal to Max | muse-wonder-1.0 | |
| Meta | Wonder 1.0 | Contributor | 1M | Minimal to Max | muse-wonder-1.0-contributor |
| Meta | Wonder 1.0 Full | 1M | Minimal to Max | muse-wonder-full-1.0 | |
| Meta | Wonder 1.0 Full | Contributor | 1M | Minimal to Max | muse-wonder-full-1.0-contributor |
| Moonshot AI | Kimi K2.8 Preview | 1M | Low to Max | kimi-for-coding | |
| Moonshot AI | Kimi K3.1 Pro | 1M | Low to Max | k3.1-pro | |
| Moonshot AI | Kimi K3.1 Pro | Highspeed | 1M | Low to Max | k3.1-pro-highspeed |
| OpenAI | GPT-6 Astra | 272K | Low to Max | gpt-6-astra | |
| OpenAI | GPT-6 Astra | Fast | 272K | Low to Max | gpt-6-astra:priority |
| OpenAI | GPT-6 Luna | 272K | Low to Max | gpt-6-luna | |
| OpenAI | GPT-6 Luna | Fast | 272K | Low to Max | gpt-6-luna:priority |
| OpenAI | GPT-6.1 Sol | 500K | Low to Max | gpt-6.1-sol | |
| OpenAI | GPT-6.1 Sol | Fast | 500K | Low to Max | gpt-6.1-sol:priority |
| OpenAI | GPT-6.1 Terra | 500K | Low to Max | gpt-6.1-terra | |
| OpenAI | GPT-6.1 Terra | Fast | 500K | Low to Max | gpt-6.1-terra:priority |
| OpenCode | OpenCode's free models, as OpenCode lists them (A free tier) | OpenCode's own, such as big-pickle | |||
| SpaceXSI | Grok 4.7 | 256K | Low to Extra high | grok-4.7 | |
| SpaceXSI | Grok 4.7 | Fast | 256K | Low to Extra high | grok-4.7-fast |
| Z.ai | GLM-5.3 | 1M | Low to Max | GLM-5.3 | |
| Z.ai | GLM-5.3 Flash | 1M | Low to Max | GLM-5.3-Flash |
Fable 5.2 starts at Medium effort on both a Claude plan and an Anthropic API key, with a 1M context window and up to 128K output tokens. Saved Fable 5.1 model selections move to Fable 5.2 within the same provider; Lookout says when the selection changes. An effort you already chose stays chosen.
Mythos 5.2 needs Anthropic's approval on each account, and it starts switched off (below). It runs once one of your signed-in accounts lists it (Lookout reads each account's model list, which costs nothing), only on the accounts that have it, and Anthropic's page in Settings › Providers says on how many of your accounts it is. Until then its row reads Off by default: needs approval for your account.
Gemini 4 Carbon, Hydrogen and Argon on Google start switched off. Lookout reads your Antigravity plan's own model list, which costs nothing: a model the plan marks unavailable reads Off by default: not on your plan, and where Lookout says why, it names the plans that have it in Google's words ("needs an Ultra plan"). A plan that includes one runs it by the name, window and effort levels the list gives. Cognition's Gemini 4 Argon is a model of its own under Cognition, with its own ID.
Choose which models your lists show
Each model under Models on a provider's page in Settings › Providers has a switch. On, the model is in your model list: the model
menu in the message box, an agent's Profile (its model, its fallbacks and the
models it may run on), the subagent model in Settings and in a project's settings, Continue on, the model a new
chat starts on, and the lists your agents read (Sessions, lookout_models, the models a Task names when it refuses).
Off, the model leaves all of them. Choose models, the settings glyph at the foot of the model menu's rail, opens that provider's page at its
Models. On Anthropic's page the Claude plan's models and the API key's each have their own group of switches. With every model switched
off, the menu says No models in your list, and a new chat still starts where it would have.
A switch hides a model; it never takes one away. A chat, project, agent, fallback list or subagent default that already uses a model you switch off keeps running on it, and where it is set the model reads Not in your model list, still there to pick. A model named by its ID, its name or its family, from a Task, the CLI, MCP or the API, runs as before. A switch never moves the model a new chat starts on from your plan to one billed per token. A model can start off, such as one that needs your account approved first: its row says so. A provider whose models are all off has no group in the model menu. The switches belong to the machine that runs your chats, as its accounts do.
Only you change them, in the app, at your terminal or from your own MCP client, as with how a provider's accounts take requests; an agent and a chat's Shell are refused:
lookout providers models PROVIDER # its models and their switches
lookout providers models PROVIDER --off ID # out of your lists
lookout providers models PROVIDER --on ID # back in
lookout providers models PROVIDER --reset ID # back to the model's defaultA switch you set for a model the provider no longer lists is kept but read nowhere; the list shows it, and --reset
clears it.
Pick a model for a chat
Each chat keeps its own model. Switch it at any time: a turn already running finishes on the model it started with, and your next message runs on the new one. Its subagents that took the chat's model move along the next time they run; one you put on a model of its own keeps it (subagents). A new chat starts on the model you picked last, while it or the same model from another provider can run. Lookout never moves a chat onto a provider billed per token, such as an API key, without you choosing it: when only such a provider could run it, the message box offers the switch instead (When a model can't run).
From a terminal, name a model by its ID:
lookout exec -m ID --effort LEVEL "PROMPT"
lookout resume THREAD -m ID "PROMPT"An ID whose provider is signed out or doesn't offer it here runs as the same model on another provider you have, and
the run says where first, as long as that provider isn't billed per token: anthropic/claude-haiku-5-5 with no
Anthropic key and a Claude plan connected runs as claude-haiku-5-5. The other way round is never automatic:
claude-haiku-5-5 with no Claude plan connected and an Anthropic key added stops before the run and names the
command that runs it on the key, billed per token (lookout exec -m anthropic/claude-haiku-5-5). An ID that can run here
always runs as named. resume with -m switches that chat to the model. --effort takes none, minimal, low, medium,
high, xhigh or max; a level the model doesn't offer moves down to the closest level it has, or up to its lowest when it
has none below (Choose the effort). Leave it out for a model whose effort is not adjustable:
Lookout refuses the run otherwise.
A model outside your plan
A model the provider still serves after your plan ended, or while it reports no plan at all, is billed per token by the provider: Meta answers Spark 1.3 and Wonder 1.0 over its paid API on the same key once a Muse Code plan has ended. Such a model ends its row in the model menu with API key, and the usage card beside the message box names it API key, Billed per token, with no meter. It runs when you pick it for a chat or an agent, and never otherwise: a new chat or agent doesn't start on it, no fallback moves to it, and an account still on the plan takes its requests first.
Choose the effort
Effort sets how long the model reasons before it answers: lower is faster, higher is smarter. The level follows the model's name in the message box, a shade lighter, and the same menu sets both: the effort sits under the models.
You can also type /effort and one of the model's levels in the message box: /effort high, /effort extra high, or a
number counting from the model's lowest level. The levels, lowest first, are None, Minimal, Low, Medium, High, Extra
high and Max. Each model offers some of them, from the lowest to the highest the Effort column above names; some skip
a level in between, and the slider shows only the model's own. When you switch models, the effort moves to the nearest
level the new model has at or below it, or to its lowest when it has none below: Max on a model that tops out at Extra
high runs at Extra high. A model whose effort is not adjustable shows its name alone, and its menu no slider.
A change of effort while a turn runs applies from that turn's next request; a change of model waits for the turn to end. The chat notes each change, who made it (you, an MCP client by its name, the fallback policy) and when it applies.
Turn on a variant
A variant is an option on the same model, such as Fast or Contributor. While it is on, the model in the message box names it, such as Spark 1.3 Contributor.
In an agent's Profile, open Model or a model under Fallbacks to set its variant there. Switching keeps the context window and any effort the selected model supports; an effort above its range moves down to a supported level.
Lookout remembers your choice for that model. A variant can be on by default: in Settings › Providers, the tooltip on the variant's badge says so. A variant can also widen the effort range: with Contributor on, Spark 1.3 adds Max.
Pick a context window
A model with more than one context window is still one row in the model menu. The chip at the end of the chat's model's row offers its windows: a Long context switch when there are two, a Context list when there are more.
When a model can't run
The model menu lists the models that can run now and that your switches show. A chat whose model can't run says why above the message box, and offers Use with the same model on a provider that can run it, named (Use Haiku 5.5 on your API key, billed per token), else with the model a new chat would start on. A switch onto a provider billed per token happens only when you click it. While the chat's stopped turn offers Continue on, the sentence drops Choose a model and Use: Continue on is the one move there. A provider whose sign-in expired, or whose plan is used up, never moves a chat to another provider on its own: sign in again, or see Accounts and usage.
| The message box says the model isn't available because | What to do |
|---|---|
| The provider isn't connected | Connect the provider. |
| The provider is signed out | Sign in to it again in Settings › Providers. |
| The provider needs a sign-in | The sign-in expired: click Sign in again on the provider's row. |
| The provider refused the sign-in | Click Sign in again on the provider's row. |
| The provider has no active plan | The plan was canceled or expired: renew it with the provider, or remove the account on the provider's row. |
| It isn't on your plan | Pick another model, or change your plan with the provider. |
| The provider refuses Lookout's client | Pick another model: its free tier stopped serving Lookout. |
| It isn't available right now | Pick another model. |
A spent plan doesn't stop you from picking a model: the menu says when its limit resets, and
Accounts and usage says what happens next. lookout exec exits with code 4 when the
model's provider needs a sign-in.