Introducing CritiqueCode All In
The coding agent. The cloud machine. The model.
One monthly price, from $10.
Write, review, and repair with CritiqueCode. All In bundles cloud sessions and inference so you can get straight to the work. Prefer standard billing? Add Inference API credits or connect your provider key for 10 hosted cloud sessions each month with no separate session fee.
Subscriptions open · No provider key required
Three monthly plans. A cloud machine and at least 2× your subscription price in included inference, measured at provider cost.
$10/month
$20 of inference + cloud sessions
Your first cloud workspace. No provider account to open.
$30/month
$60 of inference + cloud sessions
More room for regular cloud coding. The same one-bill simplicity.
$150/month
$400 of inference + cloud sessions
The frontier-model plan for founders and small teams building at full speed.
Monthly allowances reset with your billing period. On $10 and $30, the first allowance you reach—sessions or inference dollars—ends that month’s bundled usage. Founders Edition has no cloud-session cap.
The machine is included. Your inference allowance goes entirely toward model usage.
How the limits work$10 for 50 cloud sessions / month and $20 of inference · $30 for 100 cloud sessions / month and $60 of inference · $150 for unlimited cloud sessions and $400 of inference. Cloud sessions included.
Sessions and inference are tracked separately. Founders removes the session cap; its $400 inference allowance is the monthly stop.
No automatic extra charges. Upgrade from the Dashboard when you need a higher monthly allowance.
Switch models to suit the work. The announced All In lineup puts fast coding models and deeper reasoning in the same workspace.
We also offer, when capacity allows, GPT-5.6 Sol. These are additional opportunities, not guaranteed monthly access.
Use the model proxy’s highest-end routes from the same $400 monthly inference allowance—no separate provider accounts or keys.
Live availability and edition requirements appear in the session picker. Contributor edition terms are disclosed before use.
CritiqueCode is the author agent. It implements your change, then runs independent review and verified repair.
Use your own machine with BYOK or Critique inference. Your provider or your separate inference balance covers model usage.
Read the runtime contract ↗Add Inference API credits or connect an encrypted OpenRouter, LLM Gateway, or OpenAI key on /code. Both routes share 10 cloud sessions each calendar month with no separate session fee. When those sessions are used, subscribe to All In to keep working immediately.
Explore cloud Code ↗The cloud machine and inference together. No provider key and no separate model bill within your allowance.
Compare All In plans ↑The teammate that independently checks a proposed change and returns evidence. Free locally. $0.15 per cloud stage. No seats.
$15 prepaid
100 cloud stagesFor the changes that need an independent cloud finish.
$75 prepaid
500 cloud stagesFor a steady flow of changes across your team.
$300 prepaid
2,000 cloud stagesFor teams making verification part of every delivery.
Model cost stays visible. Every pack uses the same $0.15 stage rate. Model usage is always separate; local execution is free.
At your cap. Planned: an email to upgrade, with explicitly enabled continued use at $0.20/stage. Overage billing is not yet active.
The All In lineup is included. Every other hosted model stays available through prepaid model credits, with its credit floor shown in the picker.
500 credits for $10. Buy 500, 1,500, 3,000, 6,000 credit packs. Credits do not expire.
A distinct balance. Model credits extend CritiqueCode’s hosted model picker. They do not add All In sessions, refill included inference, or fund standalone Inference API calls.
For local Code, standard cloud, and standalone model usage: pay from a separate inference balance; published dollars per million token rates appear below.
Private rates shown below. Input and output are billed separately. Auto follows the model it selects. Eligible log-retention discounts remain optional.
2.5% top-up fee. Add $10 of inference for $10.25. This is standard inference pricing. A positive balance can also pay model usage inside the 10-session standard hosted Code allowance; All In top-ups are still to be announced.
| Model | Input / 1M | Output / 1M |
|---|---|---|
Critique Auto (coding)critique/auto | Routed model rate | Routed model rate |
Critique Auto Fastcritique/auto-fast | Routed model rate | Routed model rate |
Critique Auto Bestcritique/auto-best | Routed model rate | Routed model rate |
DeepSeek V4 Flash 0731deepseek/deepseek-v4-flash-0731 | $0.0650 | $0.180 |
DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash | $0.150 | $0.600 |
Qwen3.8 Flashqwen/qwen3.8-flash | $0.150 | $0.470 |
Ling 3.0 Flashinclusionai/ling-3.0-flash | $0.0750 | $0.220 |
Tencent Hy3tencent/hy3 | $0.126 | $0.522 |
NVIDIA Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b | $1.00 | $5.00 |
MiniMax M3minimax/minimax-m3 | $0.300 | $1.20 |
KAT-Coder Pro V2.5kwaipilot/kat-coder-pro-v2.5 | $0.740 | $2.96 |
Kimi K2.6moonshotai/kimi-k2.6 | $0.750 | $3.75 |
Kimi K2.7 Codemoonshotai/kimi-k2.7-code | $0.730 | $3.50 |
GLM-5.3z-ai/glm-5.3 | $1.40 | $4.40 |
Trinity Large Thinkingarcee-ai/trinity-large-thinking | $0.250 | $1.00 |
Seed 2.0 Codebytedance-seed/seed-2.0-code | $0.500 | $3.00 |
Qwen3.8 2.4T A95Bqwen/qwen3.8-2.4t-a95b | $2.00 | $6.00 |
Muse Glimmer 30Bmeta/muse-glimmer-30b | $0.350 | $1.50 |
Muse Spark 1.3meta/muse-spark-1.3 | $1.25 | $4.25 |
Qwen3.8 27Bqwen/qwen3.8-27b | $0.450 | $3.20 |
Laguna S 2.1poolside/laguna-s-2.1 | $0.0900 | $0.180 |
Grok 4.6x-ai/grok-4.6 | $2.00 | $6.00 |
Kimi K3moonshotai/kimi-k3 | $3.00 | $15.00 |
Gemini 3.8 Flashgoogle/gemini-3.8-flash | $0.750 | $3.75 |
Fugu Ultrasakana/fugu-ultra | $5.00 | $30.00 |
A cloud machine for each coding session and model inference on Critique’s account. $10/month with 50 cloud sessions / month and $20 of inference at provider cost; $30/month with 100 cloud sessions / month and $60 of inference at provider cost; $150/month with unlimited cloud sessions and $400 of inference at provider cost. Both allowances reset monthly. There is no provider key to manage and no separate model bill within your allowance.
Yes. Choose All In $10, All In $30, or All In Founders Edition on this page. Checkout is handled securely, and your plan and live balances appear in the Dashboard as soon as payment completes.
On All In $10 and $30, whichever allowance you reach first—cloud sessions or inference dollars—ends your bundled usage for that month. Founders Edition has unlimited cloud sessions, so only its inference allowance can run out. All In will not automatically charge beyond your plan. Inference API credits and provider-key users share a 10-session standard hosted allowance. When it is used, /code prompts you to subscribe to All In or wait for the next calendar month.
Yes. All In is a monthly subscription you can manage or cancel from the billing portal linked in the Dashboard. Your current allowance remains visible there.
Your included dollar allowance is measured against the provider cost of the model usage, not a count of prompts or an inflated credit value. Different models and tasks consume different amounts. Claude Sonnet 5, GPT-5.6 Terra, Kimi K3, and Qwen 3.8 Max use that allowance faster than the recommended lineup (DeepSeek V4.1 Flash, GLM-5.3 Flash, GLM-5.3). The cloud machine is included separately and does not subtract from that inference allowance.
Yes. Running the author agent on your own machine has no Critique execution fee and never uses your All In balance. You pay only for any model usage, through your own provider key or Critique inference. Local BYOK use does not require a Critique account.
All In uses Critique’s model account. On /code, you can instead connect an encrypted OpenRouter, LLM Gateway, or OpenAI key so model usage bills your provider account. Provider-key and Inference-credit sessions share 10 standard hosted sessions each calendar month with no separate session fee.
CritiqueCLI independently verifies work from the coding agent you already use. CritiqueCode is the author agent: it implements, then runs review and verified repair. Free locally. $0.15 per cloud stage; model usage separate. Starter $15 for 100 stages, Builder $75 for 500 stages, Team $300 for 2,000 stages. No seats.
The planned policy is an email prompting an upgrade at the cap. Any explicitly enabled continued cloud use would cost $0.20 per stage, above the $0.15 prepaid rate. This overage policy is not yet active; model usage remains separate.
Standard Critique inference uses a prepaid dollar balance and published input/output token rates. Top-ups add a 2.5% fee: adding $10 costs $10.25. That balance powers direct Inference API calls and can back model usage in the shared 10-session standard hosted Code allowance; it remains separate from All In and CLI stage packs.
Yes. Open /code and either add Inference API credits or connect an OpenRouter, LLM Gateway, or OpenAI key. Both routes share 10 hosted cloud sessions per calendar month at no separate session fee. Model usage bills the selected balance or provider account; after the allowance, /code prompts you to subscribe to All In or wait for the next calendar month.