Qodo pricing

About

Qodo (formerly Codium AI) is an AI code-review platform built around three surfaces presented as one product: Qodo Gen (an IDE plugin for in-editor code review and test generation), Qodo Merge (a pull-request review agent), and Qodo Command (a CLI for agentic quality workflows). As of mid-2026 the pricing page leads with code review as the core product, targeting individual developers up through large engineering organizations with distributed, multi-repo codebases.

The company markets strong adoption signals — a 4.7 rating across the VS Code and JetBrains marketplaces, over 40K weekly active users, and roughly 1,000 weekly active companies. Its differentiation centers on context-aware review across services and repositories, enforceable organization-wide quality standards, and security posture (SOC 2 Type II, two-way encryption, secrets obfuscation, and deployment options up to on-prem / air-gapped).

Qodo serves developers, team leads, and enterprise engineering leaders who want automated PR review, test generation, and policy enforcement embedded in the IDE, CLI, and Git workflow rather than bolted on after the fact. The self-serve Pro Team plan targets teams of up to 30 users; Enterprise targets 30+ users and regulated buyers.

Pricing summary : a team-wide credit pool at $.012/credit, no per-seat fee

As of June 2026 Qodo prices on pure consumption: credits at $.012 each, pooled across the whole team, with no per-seat fee and no annual commitment. Access starts with a free 14-day trial (unlimited reviews and credits, no credit card) — there is no permanent free tier. After the trial, the self-serve Pro Team plan sells selectable credit packs (the cards advertise 2,500, 5,000, and 20,000 credits, which Qodo estimates at ~18, ~36, and ~144 reviews per month) and supports up to 30 users. Enterprise is a custom plan for 30+ users, quoted via a demo.

The billing dimensions are:

• Credits — the core meter, priced at $.012/credit and pooled across the team. Every code review draws credits from the monthly pool; smaller PRs draw fewer, larger or more complex reviews draw more. A real-time dashboard shows balance and burn rate.
• Credit packs — Pro Team is sized by pack: 2,500 / 5,000 / 20,000 credits (with “see additional plan options” for more). Buyers can switch packs at any time.
• Overage — once the base pool is depleted, reviews keep running at the same per-credit rate (no premium, no penalty), accruing against a monthly spending cap the buyer sets and can change anytime.
• Premium-model multipliers — most operations cost 1 credit; premium models cost more (Claude Opus = 5 credits/request, Grok 4 = 4 credits/request), so model choice drives how fast the pool burns.
• Seats (capacity, not a fee) — Pro Team supports up to 30 users (users are not billed individually); Enterprise covers 30+ users.
• Deployment — Enterprise adds single-tenant SaaS or on-prem / air-gapped deployment plus BYOK (bring your own LLM keys).

What makes this different: Qodo has dropped both per-seat pricing and its permanent free tier, moving to a single pooled-credit balance where the team shares one meter and overage is purchasable at the base rate — the opposite of the earlier seat-plus-hidden-credit model where hitting the limit blocked work until a 30-day reset.

Pricing by product

Qodo code-review platform (self-serve plans)

Enterprise (custom plan for 30+ users)

Sales motions across products: PLG / self-serve (free trial → Pro Team credit packs) for individuals and teams up to 30 users; sales-led for Enterprise (30+ users).

Hidden costs : premium-model credit burn against a pack you sized yourself

Pro Team is priced purely on credits at $.012 each, pooled across the team — there is no per-seat fee. The cost driver is not seats but how fast the shared pool burns, and premium-model multipliers accelerate that: most requests cost 1 credit, but Claude Opus costs 5 and Grok 4 costs 4 credits per request, so a team that leans on premium models can drain a pack in a fraction of the requests a 1-credit team gets. Unlike the retired per-seat model, overage is now available at the same per-credit rate (no premium), accruing against a monthly spending cap you set — so the risk shifts from a productivity wall to an open-ended bill if the cap is set high.

A team that buys the 2,500-credit base pack (~18 reviews/month, $30 = 2,500 × $.012) and leans heavily on premium models, then runs into overage:

The hidden cost here is premium-model burn against a pack you picked yourself. A heavy Opus user exhausts the 2,500-credit base pool in roughly 500 premium requests, then keeps running at $.012/credit until the buyer-set spending cap — so the real budgeting variable is the cap, not a seat count. Sizing up to the 5,000- or 20,000-credit pack (or raising the cap) is the self-serve fix; a sales-led Enterprise quote is only needed for 30+ users or on-prem / air-gapped deployment.

Want to estimate your own Qodo bill? Use the Qodo pricing calculator to model your monthly cost based on credit-pack size, review volume, and premium-model credit burn.

Pricing evolution : from CodiumAI’s free tools to a pooled-credit consumption model

Cadence

Tracked range: 2024 Q3–2026 Q2 (qodo.ai/pricing Wayback snapshots). Earlier CodiumAI pricing pre-dates the qodo.ai domain and is reconstructed from press coverage. Quarters not listed were verified stable.

Notable changes

• 2022 — CodiumAI founded by Itamar Friedman and Dedy Kredo around test generation and code integrity.
• 2023-03 — $11M seed round; exits stealth. Open-source PR-Agent released (later Qodo Merge).
• 2024-01 — AlphaCodium open-source paper published; outperforms Google DeepMind’s AlphaCode on CodeContests (a strong community-credibility moment, 87-point Hacker News thread on the System-2/AlphaCodium write-up).
• 2024-09-30 — Rebrand to Qodo + $40M Series A (Susa Ventures, Square Peg). Teams = $19/user/mo, usage explicitly unmetered.
• 2025-03 — Teams cut to $15; Enterprise price made visible at “Starting at $45/user/mo”.
• 2025-05 — First usage meter (“messages”) layered onto the per-seat price.
• 2025-07 — Teams doubled to $30 and the meter became credit-based with premium-model multipliers; Qodo Command CLI launched.
• 2025-09 — Free-tier allowance tightened to 75 credits/month; Enterprise reverts to opaque “Contact us”.
• 2026-06-30 — Per-seat pricing and the permanent free tier both retired in one move; the $30/user/mo Teams plan and free Developer tier give way to a single self-serve Pro Team plan on a team-wide pooled-credit balance at $.012/credit (2,500 / 5,000 / 20,000-credit packs, up to 30 users, no annual commitment). Overage becomes purchasable at the same per-credit rate against a buyer-set cap, and access starts with a 14-day free trial.

The shift from unlimited to credit-metered in detail

The most consequential change is not a single price move but the mid-2025 transition from “no restrictions on calls/tokens or repositories” to a hard credit meter. At the rebrand (Sep 2024) Qodo sold a simple per-seat promise with unlimited usage. By mid-2025 it had: (1) introduced a message cap, (2) renamed it to credits, (3) made premium models cost 4–5× a normal request, (4) halved the Teams allowance while doubling the price, and (5) tightened the free tier from 250 to 75 credits. The same quarter that introduced credits doubled the Teams seat price — a rare case of a vendor raising the seat fee and adding a usage cap simultaneously, effectively repricing the product twice in one move. That seat-plus-credit phase has since been superseded: the mid-2026 repackaging dropped per-seat pricing entirely for a pooled-credit balance, and the once-”coming soon” overage path shipped — overage now runs at the same $.012/credit rate against a buyer-set spending cap, turning the meter from a hard ceiling into a revenue lever.

What’s unique : code-review-first positioning on a pooled-credit meter

1. Code review as the wedge, not autocomplete. Unlike Copilot- or Cursor-style assistants that lead with completion, Qodo’s headline surface is PR review and test generation (Qodo Merge + Qodo Gen). Pricing reflects this: packs are sized by an estimated reviews-per-month figure (~18 / ~36 / ~144), a value metric tied to review volume rather than tokens. This is a different value-metric choice than most coding-assistant peers.

2. A seat-to-pooled-credit migration that landed in 2026. Qodo spent its first year as a pure per-seat product with unlimited usage, bolted on a credit-based usage meter in 2025, then in mid-2026 dropped per-seat pricing entirely for a single pooled-credit balance shared across the team. The end state is pure consumption — the opposite migration path from vendors that started usage-metered and added seats.

3. Premium-model credit multipliers. Rather than a flat per-request charge, Qodo prices by model cost: 1 credit for standard models, 4 for Grok 4, 5 for Claude Opus. This passes foundation-model economics through to the buyer transparently, but makes a developer’s monthly capacity depend on which model they pick — an unusual coupling of model choice to plan limits.

4. A pooled, monthly-resetting balance with no rollover. Credits are shared at the workspace level and expire at the end of each monthly cycle, so a single team-wide pool — not per-user allowances — governs capacity. This makes team-level planning a matter of sizing one pack to aggregate volume, but unused credits are lost each cycle (Qodo advises picking a pack that fits typical usage rather than over-buying).

5. Deployment as the Enterprise lever. The Enterprise tier’s differentiation is less about features and more about deployment — SaaS, private cloud, on-prem, and air-gapped, with self-hosted proprietary Qodo models. For regulated buyers, deployment model (not seat count) is the primary pricing variable.

Strengths & weaknesses

Billing UX : pooled credit meter, switchable packs, and a buyer-set spending cap

• Pooled credit balance — credits at $.012 each are shared across the whole team from a single monthly pool, not allocated per seat; the docs state billing is “usage-based (not seat-based)” with “no per-user or per-engineer provisioning.”
• Selectable credit packs (“Switch packs anytime”) — Pro Team is sized by pack (2,500 / 5,000 / 20,000 credits, with “see additional plan options” for more), and buyers can upgrade, downgrade, or switch packs from the dashboard; changes take effect on the next billing cycle.
• Credit meter — every LLM request or MCP tool call inside Qodo Gen / CLI consumes credits; most operations are 1 credit, premium models cost more (Claude Opus 5 credits, Grok 4 4 credits).
• Real-time balance & burn-rate dashboard — shows current credit balance and burn rate “so you always know where you stand.”
• Overage with a self-set spending cap — once base credits are depleted, reviews continue at the same per-credit rate against a monthly spending cap you set; notifications fire as you approach the cap, and you can raise it from the dashboard so reviews never pause unexpectedly.
• Monthly reset, no rollover — credits expire at the end of each monthly cycle; each cycle starts fresh with a new pool.
• Free 14-day trial checkout (“Get Started” / “Start Now”) — sign in with GitHub / Google / email, install on repos (admin access required), reviews start with no credit card; a card is only required when picking a paid plan. No annual commitment on Pro Team.
• Enterprise demo path — a “Request a 30-Minute Executive Demo” / “Book a Demo” form for custom Enterprise (30+ users) quotes and deployment selection (single-tenant SaaS or on-prem / air-gapped, BYOK).

Strategic wins : decisions that worked in Qodo’s pricing playbook

1. Leading with code review, not autocomplete

By denominating its paid value in PR reviews per month (each credit pack is labelled with an estimated reviews-per-month figure) rather than tokens or completions, Qodo anchored its price to a metric buyers already track and value (review throughput). This sidesteps the token-anxiety problem that plagues completion-first tools and ties cost to a visible business outcome — a textbook value-metric selection. It also positions Qodo against incumbents like GitHub Copilot on a different axis (quality/governance) rather than competing head-on on completion quality.

2. A full-access trial as the PLG on-ramp

Qodo replaced its permanent free tier with a 14-day trial that grants unlimited reviews and credits with no credit card, plus a standing “Qodo for Open Source” program for qualified projects. This preserves the product-led growth funnel and open-source credibility (PR-Agent, AlphaCodium) while removing the cost of an always-free tier — a buyer experiences the full platform before the meter applies, which suits a bottom-up, developer-led motion.

3. Transparent per-model credit costs

Publishing exact credit costs per model (Opus 5, Grok 4 4, everything else 1) is unusually honest usage-based pricing communication. It lets a developer reason about capacity before committing, and it cleanly passes through the reality that premium foundation models cost the vendor more — avoiding the margin-erosion trap of flat-rate unlimited usage.

4. Deployment as the Enterprise upsell

Rather than gating core features, Qodo reserves deployment flexibility (on-prem, air-gapped, self-hosted models) for Enterprise. For regulated buyers this is the single highest-value lever, and it lets Qodo keep the self-serve tiers feature-rich while still having a compelling sales-led story. This mirrors how other enterprise AI vendors structure their gates.

Areas to improve : gaps that create friction for Qodo buyers

1. Make per-pack cost legible, not just credit counts

The page sells packs by credit count (2,500 / 5,000 / 20,000) and an estimated reviews-per-month figure, but the dollar price of each pack is only inferable from the $.012 rate (e.g. 2,500 × $.012 = $30). Concrete fix: print the explicit dollar price next to each pack so a buyer can compare cost without doing the multiplication, and show a worked overage example so the $.012 rate’s bill impact is concrete. Opaque price math invites bill-shock anxiety even when the rate itself is fair.

2. Anchor the overage spending cap with guidance

Overage now runs at the same $.012/credit rate against a buyer-set monthly cap — a real improvement over the old hard ceiling — but the page gives little help choosing the cap, so a team can over- or under-set it. Concrete fix: suggest a cap based on the chosen pack and typical premium-model mix, and surface a projected month-end spend on the dashboard, so the open-ended overage stays predictable rather than a surprise.

3. Stabilize and re-expose Enterprise pricing

Enterprise briefly carried a public “Starting at $45/user/month” anchor in early-to-mid 2025, then reverted to “Contact us.” Concrete fix: restore a public starting price or band. A visible anchor shortens sales cycles for mid-market buyers who self-disqualify when there is no number at all — a known pricing-transparency advantage.

Key takeaways

1. A pooled credit balance prices the team, not the seat. Qodo dropped per-seat pricing for a single $.012/credit pool shared across up to 30 users — the binding constraint is total credit burn, so buyers should size the pack to volume, not headcount.
2. Repackaging from seats to usage is itself a repricing event. Qodo went seat-unlimited → seat-plus-credits → pooled-credits in under two years; each move re-anchored the buyer’s mental model, and the latest one retired both the per-seat fee and the permanent free tier at once.
3. Premium-model multipliers couple capacity to model choice. A team standardizing on Claude Opus burns its pool ~5× faster than one on standard models for the same pack — model selection is effectively a budgeting decision.
4. Same-rate overage with a buyer-set cap beats a hard ceiling. Letting work continue at $.012/credit against a configurable spending cap turns the meter into a revenue lever instead of a productivity wall — the fix to Qodo’s earlier no-overage gap.
5. Deployment, not features, is the cleanest enterprise lever for regulated buyers. Qodo’s on-prem/air-gapped story justifies a custom quote without crippling the self-serve tier.

UBP implications

1. The seat-to-credit migration path is becoming common in AI coding tools. Qodo started per-seat-unlimited and added credits under cost pressure; this “seat first, meter later” pattern is the inverse of usage-native startups and signals that flat-unlimited is unsustainable once foundation-model costs bite.
2. Pass-through model pricing is a viable UBP design. Charging credits proportional to underlying model cost (1× standard, 4–5× premium) keeps margins intact while staying legible to buyers — a replicable template for any AI product reselling multiple models.
3. An overage path is table stakes for credit systems. Qodo demonstrates the failure mode of shipping a meter without a way to exceed it: the cap caps revenue and annoys power users. Usage-based pricing only works when crossing the line costs money, not productivity.

Sources

• Qodo pricing page (accessed 2026-06-30) — pooled-credit repackaging ($.012/credit, 2,500/5,000/20,000 packs, up to 30 users, 14-day trial, overage at same rate) independently re-verified 2026-06-30
• Qodo documentation: Subscription Plans (accessed 2026-06-30) — second source confirming billing is “usage-based (not seat-based)”, “no per-user or per-engineer provisioning”, credits “shared at the workspace level”, monthly renewal
• Qodo Enterprise solution page (accessed 2026-06-03)
• Qodo “Book a Demo” page (accessed 2026-06-03)
• Qodo blog: Introducing Qodo (formerly Codium) (accessed 2026-06-03)
• Qodo blog: $50M total funding announcement (accessed 2026-06-03)
• Qodo help center: Understanding Credits in Qodo (accessed 2026-06-03)
• Browse the full pricing blueprint corpus for peer AI-coding companies.

Bottom line

Qodo’s pricing tells the story of a category maturing: a free-and-unlimited CodiumAI grew into a credit-metered, seat-priced Qodo, then in mid-2026 shed per-seat pricing and its permanent free tier for a pure pooled-credit model at $.012/credit. The base 2,500-credit pack is easy to start with, but premium-model multipliers (Opus 5×, Grok 4 4×) and now-purchasable overage against a buyer-set cap are the variables buyers must model. With strong open-source credibility, an honest per-model credit table, and a deployment-led enterprise motion, Qodo is well-positioned; printing explicit per-pack dollar prices and re-exposing Enterprise pricing would remove its sharpest remaining friction points.

Compare Qodo against other AI-coding pricing models in the pricing blueprint.

Pricing timeline : Major events on a vertical axis

Each milestone below corresponds to a public pricing change, product launch, or material adjustment. Major events use a filled marker; minor adjustments use a faded one.

Repackaged to pooled credits — per-seat & free tier dropped

Jun 2026

Qodo overhauls its pricing model. The per-seat Teams plan and the permanent free Developer tier are both gone. Pricing is now a single self-serve Pro Team plan billed on a team-wide credit pool at $.012/credit, sold as selectable packs (2,500 / 5,000 / 20,000 credits ≈ 18 / 36 / 144 reviews per month), with up to 30 users, no annual commitment, and overage now available at the same per-credit rate against a spending cap the buyer sets. Access starts with a free 14-day trial (unlimited reviews/credits, no credit card) instead of a free plan; Enterprise (30+ users) stays custom-quoted with SSO/SAML, BYOK, single-tenant SaaS or on-prem/air-gapped deployment. Independently re-verified 2026-06-30 against qodo.ai/pricing and docs.qodo.ai (billing 'usage-based, not seat-based').

Per-seat snapshot — Developer / Teams / Enterprise

Jun 2026

Qodo prices its AI code-review platform on three per-seat tiers: a free Developer plan, a Teams plan shown at $30/user/month (annual billing, 'Save 21%') with a 20 PRs/user/month cap, and a contact-sales Enterprise plan adding the Qodo Command CLI, multi-repo context engine, SSO, and on-prem/air-gapped deployment. Usage inside the IDE/CLI is metered in credits (1 credit/request; premium models cost more, e.g. Claude Opus 5 credits, Grok 4 4 credits) that reset every 30 days. Teams is billed at $38/user/month month-to-month or $30/user/month on annual billing (the 'Save 21%' discount).

Free tier tightened to 75 credits; Enterprise returns to Custom

Sep 2025

Wayback snapshot shows the free Developer allowance cut from 250 to 75 credits per month (and 75 PRs/month on Qodo Merge); the previously visible Enterprise '$45/user/month' is replaced with 'Contact us'. Teams holds at $30/user/month with 2,500 credits. Supported models update to GPT-4.1/o3/o4-mini, Claude 4 Sonnet, Gemini 2.5 Pro, with Claude 4 Opus and Grok 4 Enterprise-only.

Teams doubles to $30; credits replace messages; Qodo Command launches

Jul 2025

Wayback snapshot shows Teams jump from $15 to $30/user/month while the allowance is roughly halved (5,000 messages → 2,500 credits); the meter is renamed 'credits'. Qodo Command (CLI for agentic workflows, 'in Alpha') is added across plans. Premium-model multipliers (Claude Opus 5 credits, Grok 4 4 credits) appear.

Message metering introduced

May 2025

Wayback snapshot adds usage allowances: Teams stays at $15/user/month but is now '5,000 messages', and the free Developer tier is capped at '250 messages & tool use per month'. This is the first metered usage layer on top of the per-seat price — the transition away from the earlier 'unlimited' promise.

Teams drops to $15; Enterprise price surfaces at $45

Mar 2025

Wayback snapshot shows the Teams seat price lowered to $15/user/month and, for the first time, a visible Enterprise price of 'Starting at $45/user/month'. Usage is still unmetered (no credits). AI model selection (GPT-4o, o3-mini, Claude 3.5 Sonnet, self-hosted DeepSeek-R1, Gemini 2.0 Flash) is shown across plans.

Rebrand to Qodo + $40M Series A

Sep 2024

CodiumAI rebrands to Qodo and raises a $40M Series A led by Susa Ventures and Square Peg (total funding ~$50M). The pricing page shows Developer (free), Teams at $19/user/month with a 14-day trial, and a custom Enterprise plan. The page explicitly states 'no restrictions on the number of calls/tokens or repositories' — there is no credit meter yet. Products: Qodo Gen (IDE), Qodo Merge (Git agent), Qodo Cover (CLI agent).

$11M seed, exits stealth

Mar 2023

CodiumAI exits stealth with an $11M seed round co-led by Vine Ventures and TLV Partners, with angels from OpenAI, Snyk and VMware. The flagship is Codiumate (IDE test generation); the open-source PR-Agent (later Qodo Merge) is released around this period. Pricing is free / freemium.

CodiumAI founded

Jan 1970

CodiumAI is founded by Itamar Friedman (CEO, formerly director of Alibaba's Israel AI lab) and Dedy Kredo (CPO) to build a generative-AI 'code integrity' platform focused on test generation and bug detection. The product is free during early access.