Skip to main content
ClawAI

Every AI, one workspace

Every frontier AI model in one workspace that sees, hears, researches and builds. Pay as you go, bring your team, or run it on your own hardware.

Create your accountLog inSee how it works

Start free and upgrade whenever you need more. Cancel any time.

Last reviewed 2026-07-27

See it in the app

  • A ClawAI chat with a question, the written answer, a badge naming the model that replied and the routing details under the reply.
    Every answer shows which model wrote it and why.
  • The ClawAI message box with the model picker open, listing the models you can choose from.
    Pick a model yourself, or let AUTO choose.
  • The Compare page with three models selected and a prompt typed in, ready to send.
    Ask one question and see three models answer side by side.
  • The ClawAI dashboard with an overview of your recent activity and usage.
    Your workspace at a glance.

The models you get

One workspace covers every provider below. Switch models mid-conversation, or let ClawAI pick the best one for each message.

OpenAI

50 models available now

  • GPT 5.6 Luna
  • GPT 5.6 Sol
  • GPT 5.6 Terra
  • GPT 5.5
  • GPT 5.4
  • GPT 5.4 Mini

Anthropic

13 models available now

  • Claude Opus 5.5
  • Claude Sonnet 5.5
  • Claude Fable 5.1
  • Claude Fable 5
  • Claude Opus 5
  • Claude Sonnet 5

Google Gemini

22 models available now

  • Gemini 4 Argon
  • Gemini 3.8 Flash
  • Gemini 3.7 Flash
  • Gemini 3.6 Flash
  • Gemini 3.5 Flash
  • Gemini 3.5 Flash Lite

Grok (xAI)

12 models available now

  • Grok 4.7
  • Grok 4.6
  • Grok 4.5
  • Grok 4.3
  • Grok 4.20 0309 Non Reasoning
  • Grok 4.20 0309 Reasoning

Ollama

17 models available now

  • GLM 5.3
  • GLM 5.3 Flash
  • GLM 5.2
  • Mistral Large 3:675b
  • Nemotron 3 nano:30b
  • Nemotron 3 Super

AI Horde

5 models available now

  • L3 8B Stheno v3.2
  • Gemma 4 31b
  • Gemma 4 E4B IT Ultra Uncensored Heretic
  • Gemma 4 E4B Uncensored HauhauCS Aggressive Q4 K M
  • Qwen Qwen3 0.6B IQ4 XS

Groq

4 models available now

  • GPT OSS 120B
  • Qwen3.8 27B
  • GPT OSS 20B
  • ALLaM 2 7b

Hugging Face

103 models available now

  • Kimi K2 Instruct 0905
  • DeepSeek V4 Pro 0813
  • DeepSeek V4 Flash 0731
  • DeepSeek R1 0528
  • Qwen3.5 397B A17B
  • DeepSeek V3 0324

Nvidia NIM

68 models available now

  • Palmyra Creative 122b
  • Palmyra Fin 70b 32k
  • Palmyra Med 70b 32k
  • Codellama 70b
  • llama2 70b
  • Palmyra Med 70b

OpenRouter

418 models available now

  • Mistral: Mistral Small 4
  • Mistral: Mistral Small 4 (batch)
  • Mistral: Mistral Large 3 2512 (batch)
  • Mistral: Codestral 2508
  • Mistral: Codestral 2508 (batch)
  • Mistral Large 2407

New frontier models are added as they launch — your plan covers them from day one, metered against a single allowance.

Explore every featureSee how ClawAI is built

See ClawAI in action

Short videos on what ClawAI does. Each one loads from YouTube only when you press play.

  • ClawAI — Open-Source Local-First AI Orchestration Platform
  • clawai بيجمعلك موديلات ذكااء اصطناعي في مكان واحد
  • مكان واحد في موديلات ذكاء اصطناعي كثيرة ClawAI

Plans and pricing

Every paid plan unlocks every model. The only difference is how much you can use each day and each month.

Free

Try every frontier model with a small daily allowance.

$0.00

Free trial: 90 days

Daily tokens
1,000,000
Weekly token quota
7,000,000
Monthly connector credit
No connector credit
Monthly tokens
30,000,000
Chats per day
15
Max messages per day
350
Max workspace connections
5
Max context packs
15
Max memory items
20
  • Compare mode
  • Judge mode
  • Research mode
  • Critic review
  • Workspaces
  • Memory
  • Context packs
  • Consensus Mode
  • Escalation Chain
  • Repair Lab
  • Task Decomposer
  • Best-of-N Generation
  • Verifier
  • Pipeline Lab
  • Cost-Aware Ensemble
  • Role Pack
  • Image generation & editing
  • Helper vision (image descriptions)
  • Text-to-speech
  • Plugins and MCP
Start free

Starter

Everyday access to fast, low-cost frontier models.

$4.00/month

Daily tokens
2,000,000
Weekly token quota
14,000,000
Monthly connector credit
$1.2030% of your plan price
Monthly tokens
60,000,000
Chats per day
15
Max messages per day
600
Max workspace connections
1
Max context packs
5
Max memory items
50
  • Compare mode
  • Judge mode
  • Research mode
  • Critic review
  • Workspaces
  • Memory
  • Context packs
  • Consensus Mode
  • Escalation Chain
  • Repair Lab
  • Task Decomposer
  • Best-of-N Generation
  • Verifier
  • Pipeline Lab
  • Cost-Aware Ensemble
  • Role Pack
  • Image generation & editing
  • Helper vision (image descriptions)
  • Text-to-speech
  • Plugins and MCP
Choose this plan

Plus

More allowance and access to standard-tier models.

$7.00/month

Daily tokens
3,000,000
Weekly token quota
21,000,000
Monthly connector credit
$2.1030% of your plan price
Monthly tokens
90,000,000
Chats per day
25
Max messages per day
750
Max workspace connections
2
Max context packs
15
Max memory items
200
  • Compare mode
  • Judge mode
  • Research mode
  • Critic review
  • Workspaces
  • Memory
  • Context packs
  • Consensus Mode
  • Escalation Chain
  • Repair Lab
  • Task Decomposer
  • Best-of-N Generation
  • Verifier
  • Pipeline Lab
  • Cost-Aware Ensemble
  • Role Pack
  • Image generation & editing
  • Helper vision (image descriptions)
  • Text-to-speech
  • Plugins and MCP
Choose this plan
Most popular

Pro

Premium models, deep reasoning and heavy orchestration.

$12.00/month

Daily tokens
4,000,000
Weekly token quota
28,000,000
Monthly connector credit
$3.0025% of your plan price
Monthly tokens
120,000,000
Chats per day
75
Max messages per day
950
Max workspace connections
5
Max context packs
50
Max memory items
1,000
  • Compare mode
  • Judge mode
  • Research mode
  • Critic review
  • Workspaces
  • Memory
  • Context packs
  • Consensus Mode
  • Escalation Chain
  • Repair Lab
  • Task Decomposer
  • Best-of-N Generation
  • Verifier
  • Pipeline Lab
  • Cost-Aware Ensemble
  • Role Pack
  • Image generation & editing
  • Helper vision (image descriptions)
  • Text-to-speech
  • Plugins and MCP
Choose this plan

Team

A large monthly allowance for heavy daily use.

$20.00/month

Daily tokens
5,000,000
Weekly token quota
35,000,000
Monthly connector credit
$5.0025% of your plan price
Monthly tokens
150,000,000
Chats per day
250
Max messages per day
2,500
Max workspace connections
15
Max context packs
200
Max memory items
5,000
  • Compare mode
  • Judge mode
  • Research mode
  • Critic review
  • Workspaces
  • Memory
  • Context packs
  • Consensus Mode
  • Escalation Chain
  • Repair Lab
  • Task Decomposer
  • Best-of-N Generation
  • Verifier
  • Pipeline Lab
  • Cost-Aware Ensemble
  • Role Pack
  • Image generation & editing
  • Helper vision (image descriptions)
  • Text-to-speech
  • Plugins and MCP
Choose this plan

Scale

High-volume access across every model class.

$35.00/month

Daily tokens
10,000,000
Weekly token quota
70,000,000
Monthly connector credit
$8.7525% of your plan price
Monthly tokens
300,000,000
Chats per day
1,000
Max messages per day
10,000
Max workspace connections
50
Max context packs
1,000
Max memory items
25,000
  • Compare mode
  • Judge mode
  • Research mode
  • Critic review
  • Workspaces
  • Memory
  • Context packs
  • Consensus Mode
  • Escalation Chain
  • Repair Lab
  • Task Decomposer
  • Best-of-N Generation
  • Verifier
  • Pipeline Lab
  • Cost-Aware Ensemble
  • Role Pack
  • Image generation & editing
  • Helper vision (image descriptions)
  • Text-to-speech
  • Plugins and MCP
Choose this plan

Unlimited

Unlimited chats and messages, with fair-use on premium cloud.

$120.00/month

Daily tokens
20,000,000
Weekly token quota
140,000,000
Monthly connector credit
$30.0025% of your plan price
Monthly tokens
600,000,000
Chats per day
Unlimited
Max messages per day
Unlimited
Max workspace connections
200
Max context packs
5,000
Max memory items
100,000
  • Compare mode
  • Judge mode
  • Research mode
  • Critic review
  • Workspaces
  • Memory
  • Context packs
  • Consensus Mode
  • Escalation Chain
  • Repair Lab
  • Task Decomposer
  • Best-of-N Generation
  • Verifier
  • Pipeline Lab
  • Cost-Aware Ensemble
  • Role Pack
  • Image generation & editing
  • Helper vision (image descriptions)
  • Text-to-speech
  • Plugins and MCP
Choose this plan

Cloud connectors such as OpenAI, Claude and Gemini draw from BOTH your dollar credit and your daily token allowance. Local models draw from tokens only.

Billing questions? Read the FAQSee what people build with ClawAI

Pay as you go

Pay for what the models actually use

Cloud models are metered against a credit wallet instead of being hidden inside a flat fee. A paid plan adds a monthly credit grant, top-ups never expire, and every request is reserved against your balance before it runs.

Top up only when you need to
Purchased credit never expires and stays yours through a downgrade or a cancellation — top-ups are open on the free plan too.
No surprise overspend
The cost of each request is reserved before the model runs, and a request your balance cannot cover is refused up front.
Prices in your currency
See prices in your local currency with the US-dollar amount alongside, and a ledger that shows which feature spent what.
How credit worksSee plans

What happens when you send a message

  1. You send a message in a conversation, optionally attaching files or picking a specific model.

  2. Routing picks the model based on the mode you chose, what the message asks for, and which providers are healthy right now.

  3. Context is assembled saved memories, attached context packs and the relevant chunks of your uploaded files are merged into the prompt within a token budget, alongside the conversation history.

  4. The model answers streamed back as it is generated, with live timing and reasoning where the model exposes it, and an automatic fallback to another model if the first one fails.

  5. The exchange is recorded facts and preferences worth keeping are saved for next time, and the cost is metered against your allowance so there is no surprise at the end of the month.

Memory and context are switches you control per conversation. Everything a model was given is listed in a receipt attached to the answer, so you can check what informed it. Uploaded files are chunked and retrieved the same way, which is how ClawAI answers questions about your own documents instead of guessing from training data.

Beyond a single chat window

When one model’s answer isn’t enough on its own, ClawAI offers several ways to combine, check, and improve on it.

Parallel compare
send one prompt to two to five models at once and see every response side by side, with per-model latency and token counts.
Consensus
run the same prompt across several models and synthesize a single answer from where they agree.
Escalation chains
start with a fast, cheap model and automatically escalate to a stronger one when quality falls short of a threshold.
Best-of-N
generate several candidate answers and select the strongest one by a scoring pass.
Answer repair
detect and correct specific classes of errors in a prior response rather than regenerating from scratch.
Verification
have a second pass check a response for correctness before it reaches you, with configurable revision limits.
Role packs
run a prompt through a small team of role-specialized models (for example, a planner, a critic, and a writer) that hand off to one another.
Pipelines
chain multiple orchestration stages into a single named, repeatable workflow.
Judge & Critic review
an optional independent model reviews and scores a generated response before it is treated as final.

New: ClawAI Threads

Turn a chat into a researched, cited, public article

Several models write it, a Judge and a Critic review it, and nothing goes public until you approve. Open to every plan, from Free up.

ClawAI Threads in 30 secondsA 30-second video about ClawAI Threads.30 seconds

  1. Pick a chat

    Choose “Turn into public Thread” on any chat, or start from the Threads page.

  2. Choose models and a spend cap

    Pick the authors, Judge and Critic from the same picker as chat, then set the most it may spend.

  3. Review the draft

    Read it with its sources. Ask for changes, edit it, or reject it.

  4. Approve and publish

    Only an approved draft becomes a public page. Unpublish any time.

Create a ThreadWhat Threads doesRead published Threads

Community board

See what the community has published

Browse every public Thread: researched articles, guides and explanations with their sources. Search by topic, filter by language and read for free. No account needed.

Browse the Threads board

In your editor

ClawAI works inside VS Code

Every model on your subscription, in the editor you already use. The extension is a thin client — your account, quotas and history stay on the platform, so a conversation started in the browser continues in the editor.

No second subscription
It uses the ClawAI account you already have and draws on the same allowance. Nothing extra to buy.
No API keys in your editor
Routing happens on the platform, so the extension never holds a model provider's credential.
Works with self-hosting
Point it at ClawAI's hosted platform or at your own deployment. You choose at sign-in.
Install from the MarketplaceWhat it doesInstall guide

Teams and organisations

Bring your team, keep control

Run ClawAI for a department, a company or a lab. Administrators decide who can do what, manage accounts and plans from one console, and can review what happened in the audit log.

Roles you can reshape
Every action checks a named permission, and administrators can create roles and choose what each one may do.
Accounts and plans in one console
Activate users, change roles, set temporary passwords and see each person’s usage and plan from the admin console.
An audit trail
Security-relevant actions are recorded in an audit log that administrators can filter and review.
Administration and accessTalk to us

Local-first and private

Run it on your own hardware

ClawAI is open source under the Apache-2.0 licence and runs on your own servers, with local models served through Ollama or llama.cpp and a Local-Only routing mode that keeps every request on hardware you control. For organisations, we can scope a private deployment with you.

Runs on your servers
Deployed into your data centre or private cloud, managed by your team, behind your firewall.
Local models only
Open-weight models served on your own hardware. No external provider calls and no third-party data processing.
Your controls, your audit trail
Role-based access, retention rules and an audit log of what happened, all kept inside your environment.
Contact usRead about private deployments

Start on the free plan

Create an account and send your first message in under a minute. Upgrade only when you outgrow the free allowance.

Open ClawAIView the source on GitHub