The offer, in full
Free year of Claude Team — up to five premium seats.
$1,000 in API credits — separate from the Team seats.
Eligibility: founded within the last five years, or funded within the last two.
Announced: SF Tech Week, October 2026.
What Anthropic announced
Anthropic expanded its Claude for Startups programme at SF Tech Week in October 2026. Qualifying companies get a free year of Claude Team with up to five premium seats, plus $1,000 in API credits.
The two components are independent. Team seats are for people using Claude through the web and desktop apps. The API credits are for code you write against the Claude API. A startup can lean heavily on one and barely touch the other.
The eligibility gate is unusually wide
You qualify if your company was founded within the last five years, or if you have raised funding within the last two.
That "or" is the detail worth reading twice. Most vendor credit programmes require an institutional round, which excludes bootstrapped companies entirely. Here, a four-year-old bootstrapped company with no outside investment qualifies on the founding date alone.
Worth checking before you assume you do not qualify. Solo founders, agencies, and small product companies with no venture funding routinely skip these programmes because they read "startup" as "venture-backed". On the founding-date route, neither investment nor headcount is tested.
What five Team seats are actually worth
Claude Team is Anthropic's collaboration tier — shared projects, higher usage limits than individual plans, and central billing. Five premium seats for twelve months is a substantial amount of capacity for a small company, and it is the part of this offer most people will use daily.
If your team is larger than five, the programme covers five seats and you pay for the rest. It does not scale with headcount.
What $1,000 in API credits buys
This depends entirely on which model you call, and the spread is enormous. Using Anthropic's published Claude Haiku 5.5 pricing of $0.10 per million input tokens and $0.50 per million output tokens for prompts up to 100K tokens, $1,000 covers a very large volume of lightweight work.
Call a frontier model instead and the same $1,000 goes a fraction as far. The practical advice is simple: route everything you can to the cheapest model that does the job, and spend the credits where the capability gap is real.
We covered the model economics in detail in Claude Haiku 5.5 pricing and benchmarks, and the free-tier boundaries in Claude free plan limits and API access.
Where the limits bite
- It is a year, not a permanent tier. Build your cost model for month thirteen now, while it is cheap to change your mind.
- Five seats is a hard cap. A growing team will outrun it.
- $1,000 is not a lot of frontier-model inference. It is generous for prototyping and thin for production at scale.
- Credits expire. Vendor credits typically carry an expiry; confirm yours at application rather than discovering it in month eleven.
How to decide whether it is worth applying
If you are already paying for Claude Team — apply immediately. You are converting a line item into zero for a year.
If you are prototyping with the API — apply. $1,000 is more than most prototypes consume before they either work or get killed.
If you are already running production inference at volume — apply, but do not restructure anything around it. A thousand dollars will not change your unit economics.
If you have not used Claude at all — the Team seats matter more than the credits. Start there.
How this compares to other vendor credits
Credit programmes from the major labs and clouds move constantly, and published figures go stale within months. Two things are durable enough to plan around.
First, the wider eligibility here is genuinely unusual, and that is the differentiator rather than the dollar amount. Second, credits are a customer-acquisition cost, not a discount — they are priced so that the year of free usage builds the integration that makes year two hard to leave. That is not a criticism, it is simply the trade. Take it with open eyes and keep your abstraction layer clean.
FAQ
Who qualifies for Claude for Startups?
Companies founded within the last five years, or companies that have raised funding within the last two. Meeting either condition is sufficient.
What exactly do you get?
A free year of Claude Team with up to five premium seats, plus $1,000 in API credits.
Do bootstrapped companies qualify?
On the founding-date route, yes — the five-year condition does not test for outside investment.
Is the $1,000 separate from the Team seats?
Yes. The API credits are for programmatic use through the Claude API; the Team seats are for people using Claude through the apps.
What happens after the free year?
Anthropic has not announced any automatic continuation. Plan for standard Claude Team pricing from month thirteen.
Can I use the credits on any Claude model?
API credits apply to API usage. How far they go depends heavily on model choice — lightweight models stretch them dramatically further than frontier models.