04 / Pricing

Pay for what
you actually use.

No flat PR limits. No hidden model routing. Every plan gives you a monthly credit pool — one review typically uses 3–50+ credits depending on model and PR size, so your pool covers hundreds of reviews, not raw credit count as jobs.

Credits explained

One system.
Simple rules.

Lead Model
The main reviewer. Reads the PR, synthesises all findings, writes the final verdict. Consumes the majority of credits — the brain of every review.
Sub-Agents
Focused specialists — security, test gaps, architecture, performance. Each adds a small credit overhead on top of the lead model cost.
PR Depth
Tiny PRs cost less than large ones. A depth multiplier scales the total based on changed files, context size, and whether you're re-running or escalating.
Lead Model
credits
+
All Sub-Agents
credits
×
Depth Multiplier
1.0×
=
Total PR Cost
credits
PR SizeChanged FilesMultiplierExample: GLM-5 + 2× GLM-4.7 Flash (~4.5 base cr)
Tiny1–50.75×~3 credits
Normal6–151.0×5 credits
Large16–301.5×~7 credits
Very Large30+2.0×9 credits
Pricing
Choose a plan

Three tiers.
One philosophy.

Standard
$12/mo
billed monthly
500 credits / month
Use efficient models and cover a lot of ground.
Get started
ProRecommended
$35/mo
billed monthly
7-day free trial
2,000 credits / month
Balance volume with premium reasoning where it matters.
Start free trial
Ultra
$129/mo
billed monthly
10,000 credits / month
Frontier models plus any lead as sub-agent (e.g. GPT-5.4 Pro lead + Claude Sonnet 4.6 sub). No routine limits.
Sign up
EnterpriseCustom solutions

Scale with custom volume, SLAs, and deployment options. On‑prem or cloud, with dedicated support and full compliance.

Custom volume & SLAsLocal / on‑prem hostingDedicated supportSSO & compliance
Sign up
Pricing
Model credit guide

Every model.
Every cost.

Tier:
Model
Credits / review
Type
Plan
Everyday volume — Lead models
DeepSeek
DeepSeek V3.2
DeepSeek
1
Lead
Standard+
Minimax
MiniMax M2.5
MiniMax
2
Lead
Standard+
Kimi
Kimi K2.5
MoonshotAI
3
Lead
Standard+
Z.ai
GLM-5
Z.AI
3
Lead
Standard+
Google
Gemini 3 Flash
Google
4
Lead
Standard+
Qwen
Qwen3.5-397B-A17B
Alibaba
6
Lead
Standard+
Standard — Specialist sub-agents
Z.ai
GLM-4.7 Flash
Z.AI
0.75
Sub-agent
Standard+
Qwen
Qwen3 Coder Next
Alibaba
1
Sub-agent
Standard+
KwaiKAT
KAT-Coder-Pro
KwaiPilot
1
Sub-agent
Standard+
Google
Gemini 3.1 Flash Lite
Google
2
Sub-agent
Standard+
OpenAI
GPT-5.1 Codex Mini
OpenAI
3
Sub-agent
Standard+
Pro — Premium lead models
$35/m
OpenAI
GPT-5.1 Codex
OpenAI
12
Lead
Pro+
Google
Gemini 3.1 Pro
Google
16
Lead
Pro+
OpenAI
GPT-5.3 Codex
OpenAI
16
Lead
Pro+
OpenAI
GPT-5.4
OpenAI
20
Lead
Pro+
Claude
Claude Sonnet 4.6
Anthropic
22
Lead
Pro+
Pro — Premium sub-agents
Mistral
Devstral 2512
Mistral
3
Sub-agent
Pro+
Stepfun
Step 3.5 Flash
StepFun
1
Sub-agent
Pro+
XiaomiMiMo
MiMo V2 Flash
Xiaomi
1
Sub-agent
Pro+
Inception
Mercury 2
Inception
1
Sub-agent
Pro+
Qwen
Qwen3.5-122B-A10B
Alibaba
4
Sub-agent
Pro+
Ultra — Frontier models
Claude
Claude Opus 4.6
Anthropic
37
Ultra
Ultra
OpenAI
GPT-5.4 Pro
OpenAI
241
Ultra
Ultra
Ultra — Any lead model as sub-agent

On Ultra, any lead model (including Pro+ e.g. Claude Sonnet 4.6) can be used as a sub-agent. Example: GPT-5.4 Pro (lead) + Claude Sonnet 4.6 (sub). Standard and Pro: sub-agents are from the fixed list only — you cannot use Sonnet or other lead models as sub-agents.

Credits are calculated per review unit — approximately 100k input + 15k output tokens for the lead model on a normal PR.

Pricing
How far do your credits go?

Pick your stack.
See the math.

Select a lead model, a specialist sub-agent, the number of sub-agents, and your typical PR size. We calculate the real credit cost and show how many reviews each plan covers.

Lead Model
Specialist Sub-Agent
Number of sub-agents
PR size / depth
(1 lead + 0.75 × 1 sub) × 1× =2 credits / PR
Standard
$12/mo
250
PR reviews / month
500 credits ÷ 2 credits
Pro
$35/mo
1,000
PR reviews / month
2,000 credits ÷ 2 credits
Ultra
$129/mo
5,000
PR reviews / month
10,000 credits ÷ 2 credits
Pricing FAQ
01

What counts as one credit?

1 credit = roughly 100k input + 15k output tokens for the lead model on a normal PR. Sub-agent costs add on top; depth multiplies the total. So one full review often uses 3–50+ credits depending on model choice and PR size — your monthly pool covers hundreds of reviews, not raw credit count as "jobs".

02

What's the difference between Standard, Pro, and Ultra?

Standard gives you efficient lead models and a fixed list of specialist sub-agents. Pro adds premium lead models (e.g. GPT-5.4, Claude Sonnet) and more sub-agents — but sub-agents are still from a fixed list only. Ultra adds frontier leads (Claude Opus, GPT-5.4 Pro) and lets you use any lead model as a sub-agent (e.g. GPT-5.4 Pro as lead + Claude Sonnet 4.6 as sub), which other tiers cannot do.

03

Can I upgrade or downgrade mid-month?

Yes. Upgrades take effect immediately; downgrades apply at the start of your next billing cycle.

04

Do unused credits roll over?

Unused monthly credits do not roll over. The pool resets at the start of each billing period.

Ready to ship better code?

Start reviewing in minutes.

Connect your GitHub repo, pick a plan, and let Critique handle the rest.