AWS supports two distinct paths for issuing startup credits: the self-serve Activate forms at aws.amazon.com/startups/credits, and the partner-filed route where a vetted AWS partner submits via the ACE program on the startup’s behalf. Different ceilings, different timelines, different eligibility floors, different reviewer trust. This page walks through every dimension and the situations where each is the right move.
Before comparing dimensions, the two paths need to be cleanly defined. They are not two ways of filing the same form — they are two structurally different routes that produce different credit amounts via different reviewer pools at different speeds.
The self-serve path runs through the public AWS Activate console at aws.amazon.com/startups/credits. Any incorporated company (LLC, C-Corp, or international equivalent) can open the page, click into the Activate Founders application, fill in a short form, and submit. The form asks for company name, company URL, country of incorporation, AWS account ID, and a 3–5 sentence use case description. There is no investor reference, no projected-spend itemization, no partner involvement. The reviewer is an AWS-internal screening team (largely automated with human spot-check) that pattern-matches against eligibility criteria and either approves the $5K Founders award or downgrades to the $1K Builders tier.
The partner-filed path runs through a different system entirely. The startup never touches the Activate form. Instead, a vetted AWS partner — an agency, consultancy, or solutions integrator with Advanced or Premier tier in the AWS Partner Network — opens the ACE portal (APN Customer Engagements), creates an opportunity record describing the startup, the use case, the projected AWS consumption itemized by service, and the engagement type, and submits the record for review. The reviewer on the partner-filed side is a partner-development manager (PDM) inside AWS who reads the record, applies an approval discretion in the relevant credit-pool range, and posts the credit to the startup’s AWS billing console.
These are two genuinely separate AWS systems. The self-serve form does not roll up into ACE; ACE submissions do not appear in the public Activate console as pending applications. The reviewer pools do not overlap. The credit pools do not overlap (Founders self-serve is one pool; partner-filed Portfolio is a different pool with a separate budget). The decision of which path to file is, mechanically, a decision about which system to enter — not a decision about which form within one system to fill out.
This separation matters because it determines which dimensions you should compare. If both paths fed into one queue with one reviewer, the comparison would collapse to “how much credit do you ask for.” Because they are separate systems, the comparison runs across six independent dimensions, each with different trade-offs.
Most “Activate vs partner-filed” discussions collapse to ceiling — “self-serve is $5K, partner-filed is $100K, pick partner-filed if you qualify.” That collapses too far. The honest comparison runs across six dimensions, and at least two of them frequently push toward self-serve even when partner-filed is technically available.
The six dimensions, each examined in detail below: ceiling, application time, approval window, eligibility floor, reviewer trust, and future stack-friendliness. The first three are quantitative; the second three are qualitative but consequential.
Self-serve ceiling: $5K via Activate Founders. The $1K Builders tier sits below as the unincorporated alternative. There is no self-serve form above $5K. The Activate console does not expose Portfolio, Build for Startups, or Bedrock POC as self-serve options because AWS does not want $50K–$100K credit pools issued without partner-attested verification.
Partner-filed ceiling: $25K for Founders via ACE (bootstrapped-eligible), $50K–$100K for Portfolio Sub-Program (institutionally-vouched), plus stack-additives reaching $150K total when Build for Startups (+$25K) and Bedrock POC (+$25K typical, up to +$50K ceiling) layer on a Portfolio base. The partner-filed ceiling is a function of which pools the partner files, not a single number.
The 30× ratio between self-serve $5K and partner-filed Portfolio $100K-plus-additives toward $150K is the headline number, and the reason most founders default to wanting partner-filed. But ceiling alone is not the deciding factor — the next five dimensions modify the picture.
Self-serve application time: ~5 minutes. The Activate Founders form is short: 4 fields and a use case paragraph. A founder who has the AWS account ID handy can complete the form between two meetings without prep.
Partner-filed application time: ~30 minutes of founder time plus 1–3 hours of partner time. The partner runs a discovery call (~30 minutes), then the founder fills in a worksheet (company info, AWS account ID, deck, use case paragraph, projected spend by service for the next 12 months, evidence of any additive workloads). The partner takes the worksheet inputs and assembles the ACE record (description, projection, engagement scope, funding source). Founder net time across the whole partner-filed application: between 25 and 35 minutes depending on how thorough the worksheet is.
The application-time gap (5 minutes vs 30 minutes) is real but small in absolute terms. The 25-minute incremental cost in founder time translates to roughly $1,000 per minute when the partner-filed path lands $25K incremental credits, and roughly $3,200 per minute when the path lands $95K incremental (Portfolio after a Founders downgrade). The dollar-per-minute math overwhelmingly favors partner-filed whenever the founder qualifies.
Self-serve approval window: 3–7 days from form submission. The automated screening usually completes within 24–48 hours; the human spot-check and credit posting typically take another 1–5 days. Well-formed applications with a clean entity and a plausible use case approve at the fast end of the window; applications with marginal use cases or ambiguous entity records sit closer to the 7-day end.
Partner-filed approval window: 10–18 days from ACE submission. The partner-development manager review is queued behind other PDM work; the discretionary credit assignment requires more reviewer time than the self-serve screening; the credit posting flow runs through a different internal team than the Activate console. Founders applications via ACE approve closer to the 10-day end; Portfolio applications closer to the 14–18-day end because the dollar amount triggers additional review.
The 7–11-day gap in approval window is the most consequential single number when the credits are time-sensitive. A founder who needs credits to spin up infrastructure for a customer demo next week does not have 18 days. A founder building toward a 6-month runway has plenty of time for partner-filed. The “$5K bridge” pattern (Section IV) exists precisely because some founders need both timelines simultaneously.
Self-serve eligibility floor: any AWS-eligible incorporated company. The bar is deliberately low — incorporated entity, AWS account in the company name, plausible AWS use case, non-sanctioned region, company age under 10 years. No funding stage requirement. No revenue requirement. No projected-spend threshold.
Partner-filed Founders eligibility floor: no funding-stage requirement. Bootstrapped startups, revenue-funded startups, pre-seed startups with no institutional capital all qualify for the partner-filed Founders track at up to $25K. The partner attests to the workload being real and projectable; the partner’s ACE track record substitutes for the institutional vouch.
Partner-filed Portfolio eligibility floor: institutional vouch required. The Portfolio Sub-Program is gated to startups with one of (a) a venture-capital investor with Portfolio Sub-Program access who can file directly, (b) a partner submitting via ACE who can demonstrate institutional backing in the engagement record, or (c) an accelerator with Portfolio batch access (top-tier accelerators only). Pre-seed startups with no institutional vouch are not eligible for Portfolio and will be downgraded to Founders if a partner files Portfolio on their behalf without the vouch.
Self-serve reviewer trust: self-attested. The applicant fills in their own description of the use case and the company. The reviewer cannot verify projected AWS consumption (no detailed projection is requested) or verify operational maturity (no deck is requested). The reviewer can verify the entity exists (registry lookup) and the use case is plausible (pattern-matching). Self-attested submissions are capped at the $5K Founders ceiling because the verification depth supports only a small dollar award.
Partner-filed reviewer trust: partner-attested. The AWS partner has skin in the ACE program — partners with high ACE close rates and clean track records receive fast-track treatment; partners with rejection histories or poor record quality see their submissions scrutinized more heavily. The partner attesting to the use case, the projection, and the engagement scope shifts verification burden from AWS reviewers to the partner. This is the mechanic that lets the partner-filed ceiling go to $100K and beyond — the trust transfer makes higher dollar awards viable.
The trust dimension explains why AWS will not lift the self-serve ceiling above $5K and why the partner-filed mechanic exists at all. AWS could theoretically expose a Portfolio self-serve form; they choose not to because the rejection cost of bad approvals at $100K dramatically exceeds the verification cost of routing through a vetted partner.
Self-serve stack-friendliness: high. Claiming the $5K Founders self-serve award does not block any future partner-filed application. A founder who self-serves $5K on day 1 can return six months later and file partner-filed Founders for up to $20K incremental (reaching the $25K Founders ceiling), or Portfolio for up to $95K incremental (reaching the $100K Portfolio ceiling). The Bedrock POC track and Build for Startups track stack independently and are not reduced by the prior self-serve award.
Partner-filed Portfolio stack-friendliness: medium. Filing partner-filed Portfolio uses up the Portfolio slot for the 24-month credit validity window. A startup that lands $50K of Portfolio cannot return three months later and file another Portfolio application for additional dollars under the same legal entity — the slot is consumed. Stack-additives (Build for Startups, Bedrock POC) remain available because they are separate pools, but the Portfolio pool itself is a single-issuance-per-entity-per-validity-window resource.
The stack-friendliness asymmetry is one reason the $5K bridge pattern works: self-serve does not consume any of the partner-filed slots, so filing self-serve first is strictly additive. The reverse is not true — filing partner-filed first does consume the Portfolio slot and changes the future application math.
A common founder mistake is treating the two paths as mutually exclusive — “I should either go self-serve or partner-filed, not both.” The opposite is true for many profiles. The $5K bridge pattern is filing both in parallel, with the self-serve credits used as a bridge during the partner-filed review window.
The mechanics of the bridge: on day 1, the founder submits the self-serve Activate Founders form (~5 minutes of work). On day 1 or day 2, the founder kicks off the partner-filed application — a CloudRoute inquiry routes to a vetted partner within 24 hours, and the partner runs a discovery call within the following 48 hours. The partner files the ACE record around day 4 or day 5. By day 4–7, the self-serve $5K has landed in the AWS billing console; the partner-filed Portfolio is in review. The startup uses the $5K to spin up infrastructure, run experiments, or fund the SOC 2 remediation work the partner is delivering alongside the credit application. Around day 14–18, the partner-filed Portfolio lands ($50K–$100K), often with Build for Startups (+$25K) and Bedrock POC (+$25K) stack-additives layered on. Total credit position at the end of the bridge: $80K–$155K depending on which additives apply.
The bridge is most useful for three profiles. First: institutionally-funded startups with an immediate infrastructure need — a customer demo, a compliance audit, a migration kickoff — that cannot wait 14 days. The bridge converts the partner-filed wait into productive infrastructure time. Second: bootstrapped startups testing the partner-filed track for the first time and uncertain whether they will qualify for Portfolio (vs being downgraded to Founders). The bridge guarantees at least $5K of credits regardless of the partner-filed outcome. Third: founders who want to start AWS experimentation immediately while the partner-filed application proceeds in parallel.
The bridge does have a small caveat. The self-serve $5K is the Activate Founders pool; the partner-filed Portfolio is the Portfolio pool. These are different budgets and different reviewer pools, so they do not interfere with each other procedurally. But the Portfolio award absorbs the Founders sub-program — meaning when Portfolio lands, the $5K from self-serve does not stack mathematically as $5K + $100K = $105K on top of Portfolio’s own internal Founders coverage. Portfolio includes Founders-level coverage by design. The bridge value is the time bridge (using the $5K during the partner-filed wait), not arithmetic addition. Section VII covers this stacking nuance in detail.
For most institutionally-funded applicants pursuing Portfolio, the recommended sequence is: file self-serve immediately, kick off the partner-filed process within 48 hours, plan around credit availability at day 4–7 (small $5K window) and day 14–18 (full $100K-plus-additives window). The combined founder time across both paths is ~35 minutes; the combined credit landing is between $50K (downgrade scenario) and $155K (full stack scenario).
The self-serve path is the correct ask in a specific set of situations. These are the profiles where the partner-filed mechanic would be over-engineered, where the dollar gap is small enough that the time gap dominates, or where the eligibility floor for partner-filed Portfolio is not met.
The mirror of the previous section. These are the profiles where the partner-filed path is unambiguously the right ask, and where defaulting to self-serve leaves substantial credit dollars on the table.
The two paths interact asymmetrically with eligibility. Picking the wrong one is not free. Two specific downgrade scenarios are worth understanding because they are the most common ways founders leave credit dollars on the table.
Downgrade scenario 1 — applying for self-serve when partner-filed Portfolio was available. A Series-A startup with a tier-1 VC files the self-serve Activate Founders form on day 1, gets approved for $5K on day 4, and considers the credit question closed. The startup has actually qualified for Portfolio Sub-Program but never filed for it. The opportunity cost: $95K of credits ($100K Portfolio ceiling minus the $5K already received). The Portfolio path is not blocked by the prior self-serve issuance — the startup could still file partner-filed Portfolio months later — but in practice, most startups do not revisit the credit question once they have a credit balance. The $5K self-serve “closes” the mental loop; the $95K of foregone credits never gets recovered.
Downgrade scenario 2 — applying for partner-filed Portfolio when the startup is ineligible. A pre-seed bootstrapped startup with no institutional vouch (no VC, no top-tier accelerator) engages a partner to file Portfolio. The AWS reviewer reads the ACE record, notes the absence of institutional backing, and downgrades the award to Founders ($25K typical). This is not technically a rejection — the startup still receives $25K — but the founder went into the engagement expecting $100K and lands at $25K. The 4× shortfall is real. Worse, in some cases the reviewer flags the record for “overreach” and the partner’s ACE track record takes a hit, which affects the partner’s standing on future submissions.
The right framing for these scenarios: the two paths have different qualification floors. Self-serve has a low floor and a low ceiling. Partner-filed Founders has a low-to-mid floor and a mid ceiling. Partner-filed Portfolio has a higher floor (institutional vouch) and the highest ceiling. Choosing the path is not just about picking the bigger number — it is about matching the path to the actual qualification profile.
CloudRoute’s triage on inbound inquiries is essentially a pre-check on this matching. The triage asks: what is the institutional standing? what is the AWS use case? what is the projected burn? what stack-additives apply (AI, migration)? The output of triage is which path (or combination of paths) is the highest-expected-credit move for this specific startup. The pre-check exists because the downgrade scenarios are real and asymmetric — filing the wrong path either leaves money behind (self-serve when Portfolio was available) or sets the wrong expectation (Portfolio when only Founders was achievable).
For founders reading this without CloudRoute triage: the decision tree at the end of Section IX captures the core logic. The short version is — if any institutional vouch is present, go partner-filed Portfolio. If no vouch but a real workload exists, go partner-filed Founders. If neither vouch nor projectable workload is present, go self-serve.
Y Combinator companies are a special case in this comparison because the self-serve path is partially automated for them. Understanding what YC standing covers (and what it does not) is important for YC companies and for anyone benchmarking against the YC credit position.
The YC $5K credit is the same Activate Founders pool the public self-serve form awards. It is not a separate YC-only allocation. AWS maintains a batch integration with YC that auto-issues the Founders credit to every accepted company at batch start; the credit shows up in the AWS Activate dashboard under the relevant YC batch designation (e.g., “Y Combinator W26”). The founder does not fill in the public Activate Founders form — the credit is already issued.
The mechanical implication for YC companies pursuing higher credits: the self-serve Founders slot is already consumed by the YC auto-issuance. The YC company should not re-file the public form (it will be denied as a duplicate). The YC company should pursue partner-filed Portfolio for the additional $95K — the YC standing does not block Portfolio access. In fact, YC standing strengthens the Portfolio application because YC counts as a top-tier accelerator institutional vouch for Portfolio eligibility purposes.
The cumulative credit position for a YC company doing both paths: $5K (YC-issued via the auto integration) + $50K–$100K (partner-filed Portfolio) + stack-additives ($25K Build for Startups, $25K Bedrock POC) = potentially $155K total. The $5K does not stack arithmetically with Portfolio (Portfolio absorbs Founders-level coverage) but the $5K does land in the AWS billing console before the Portfolio lands, providing the same bridge value as the explicit $5K-bridge pattern described in Section III.
Techstars, 500 Global, Antler, MISK, Flat6Labs, MassChallenge, and most major regional accelerators have similar batch integrations. The credit amounts vary by accelerator — Techstars typically issues $5K matching the Founders tier; some smaller accelerators issue $1K Builders-tier credits; some larger ones issue $10K through partner relationships. The check is always the same: log into the AWS Activate console, look at promotional credits, see what is already issued, plan additional applications around what is missing.
A subtle but consequential point about the two paths: stacking them does not work the way pure arithmetic would suggest. The Portfolio award absorbs Founders-level coverage; Build for Startups and Bedrock POC genuinely stack on top. Getting this wrong leads to founder expectation mismatches.
The non-stacking case: self-serve $5K + partner-filed Portfolio $100K does NOT equal $105K. Portfolio is a higher-tier pool that absorbs Founders-tier coverage by design. When Portfolio lands at $100K, the $5K self-serve award does not stack on top as a separate balance — the Portfolio award becomes the operative credit pool, and the $5K (if still present in the billing console with remaining balance) is consumed alongside but does not extend the ceiling beyond $100K.
The practical exception: timing. If the self-serve $5K landed in day 4 and the founder spent $3K against it before Portfolio landed in day 16, the founder effectively used $3K of credits during the bridge window, plus the full $100K of Portfolio when it landed. Net consumed credit: $103K of credit value over the lifecycle, even though no arithmetic sum of awarded amounts exceeded $100K at any single point. This is the time-bridge value of running self-serve in parallel with Portfolio.
The genuine stacking cases: Build for Startups and Bedrock POC. Both are separate pools with separate review tracks and separate budgets. Build for Startups adds up to $25K on top of Founders or Portfolio when a distinct migration or build workload is attested. Bedrock POC adds $10K–$50K on top of either base when a distinct Bedrock-based AI workload is attested. The reviewer for each of these tracks is checking that the workload is genuinely additive (not double-counting against the base pool), but when the workload qualifies, the additive amount stacks arithmetically.
The realistic stacking ceiling: $150K at Series-A. That is $100K Portfolio + $25K Build for Startups + $25K Bedrock POC. Outlier cases reach $200K when there is a substantial migration that escalates the Build for Startups award beyond the typical $25K (this enters MAP territory) or a Bedrock POC awarded at the $50K ceiling instead of the $25K typical. The $150K number is the working ceiling for planning purposes; the $200K cases require additional partner-filed work and are not the default outcome.
For startups planning the full stack, the partner runs separate ACE records — one for Portfolio, one for Build for Startups, one for Bedrock POC. All three submitted in the same week typically land within the same 14–18 day window. Filing them at staggered times does not change the landing time materially; filing them at the same time keeps the founder’s involvement coordinated to one set of worksheets and one decision cycle.
Reduced to operational form, the decision between self-serve and partner-filed (and combinations of the two) comes down to five sequential questions. Answer them in order; the path falls out.
The full comparison table across the six dimensions covered in Section II plus validity and cost. Self-serve dominates on speed and floor; partner-filed dominates on dollars and additives. Cost is $0 on both because AWS funds the credit pools and (in the partner-filed case) the partner is paid by AWS through engagement-funding programs separate from the customer.
| Variable | Activate self-serve ($5K Founders) | Partner-filed ($25K–$100K + stack) |
|---|---|---|
| Ceiling | $5K (Founders) or $1K (Builders) | $25K (Founders) / $100K (Portfolio) + up to $50K additives → $150K |
| Application time (founder) | ~5 minutes | ~30 minutes (worksheet + discovery call) |
| Approval window | 3–7 days | 10–18 days (Founders fast end, Portfolio slow end) |
| Eligibility floor | Any incorporated AWS-eligible LLC | Founders: no funding stage. Portfolio: institutional vouch required. |
| Reviewer trust | Self-attested (automated screening + spot-check) | Partner-attested (PDM discretion, partner ACE track record) |
| Validity | 12 months | 24 months (Portfolio) / 12 months (Founders, Build for Startups, Bedrock POC) |
| Stack-additives | None (self-serve does not expose Build / Bedrock pools) | Build for Startups +$25K, Bedrock POC +$10K–$50K |
| Cost to customer | $0 | $0 (partner paid by AWS engagement-funding, separate from customer credits) |
Situation: Already on AWS for 11 months. Needed credits to fund (a) an immediate infrastructure spike for an enterprise customer pilot scheduled to start in 5 days, (b) SOC 2 Type 2 remediation work spanning the next 8 weeks, (c) a Bedrock-based AI assistant feature targeted for production in 12 weeks. Internal infra lead was at 80% allocation on the customer pilot work and could not take on the SOC 2 or Bedrock work without external help. Founder did not want to pay out of pocket for any of the three scopes.
What CloudRoute did: Day 1 — founder submitted self-serve Activate Founders application for $5K bridge credits. Day 1 — founder also submitted CloudRoute inquiry; routed within 18 hours to a US-West partner with SOC 2 + Bedrock + B2B SaaS track record. Day 3 — discovery call with partner. Day 4 — $5K self-serve credit landed in the billing console; founder used $1,200 of it that day to spin up the enterprise pilot infrastructure. Day 5 — partner submitted three ACE records: Portfolio ($100K base for general AWS infrastructure), Build for Startups ($25K for the SOC 2 remediation engagement), Bedrock POC ($25K for the AI assistant feature). All three within the same week.
Outcome: Day 18 — all three partner-filed awards approved at the requested amounts. Total credits in the AWS billing console: $5K bridge (with ~$3,800 remaining balance at that point) + $100K Portfolio + $25K Build for Startups + $25K Bedrock POC = $155K aggregate credit position including the bridge. SOC 2 remediation completed in 7 weeks (fully credit-funded). Bedrock assistant in production at week 11. Enterprise pilot landed without infrastructure friction. CloudRoute’s commission was paid by the partner from AWS’s engagement funding. Founder paid $0 across all four awards.
engagement window: 18 days · founder time: ~38 minutes across both paths · credits secured: $155K · cost to customer: $0
If your situation fits self-serve, the public Activate form is the fastest path. If you might qualify for partner-filed Portfolio or stack-additives, CloudRoute routes you to a vetted partner who files the ACE record. Customer pays $0 on either path.