fullstack-builder / rank 33 of 41 / cycle 04
Caffeine benchmark
The only product in the index that deploys to a decentralised runtime, which makes it the hardest to compare fairly. Data durability and tenancy were strong, and the application kept running with no server to maintain. Everything that assumes a conventional web stack, from the crawlable page test to the outbound webhook prompt, either scored low or did not apply, and the first paint numbers are the worst in the cohort.
Composite index
60.10
of 100 | ▲ 1 versus cycle 03
At a glance
Composite index
60.10
rank 33 of 41
Rank
#33
of 41 measured builders
Median speed
150.6 s
p10 76.0 s, p90 295.5 s
Reliability
66.7%
strict pass rate
SEO score
48.0
0 to 100, SEO and GEO axis
Three strongest axes
- Value76.0
- Scalability74.0
- Agent performance70.0
Three weakest axes
- Speed20.8
- Integrations46.0
- SEO and GEO48.0
Median per prompt
150.6 s
p10 76.0 s, p90 295.5 s
Strict pass rate
66.7%
30 pass, 12 partial, 3 fail
Executions recorded
45
5 runs of 9 prompts
Entry price
EUR 0.00
Cheapest plan that can finish the specification
Per axis subscores
Composite recomputed from these ten rows: 60.10. That is the arithmetic behind the index, and you can repeat it by hand.
| Axis | Subscore 0 to 100, tick is the roster median | Score | Median | Weight % | Contribution |
|---|---|---|---|---|---|
| Agent performance | 70.0 | 75.0 | 18 | 12.60 | |
| Reliability | 66.7 | 71.1 | 18 | 12.01 | |
| Scalability | 74.0 | 70.0 | 13 | 9.62 | |
| SEO and GEO | 48.0 | 58.0 | 12 | 5.76 | |
| API and MCP | 52.0 | 62.0 | 9 | 4.68 | |
| Integrations | 46.0 | 66.0 | 8 | 3.68 | |
| Design output | 68.0 | 71.0 | 7 | 4.76 | |
| Speed | 20.8 | 33.0 | 7 | 1.46 | |
| Value | 76.0 | 76.0 | 5 | 3.80 | |
| Code ownership | 58.0 | 90.0 | 3 | 1.74 |
Speed distribution
Every recorded prompt duration for Caffeine, in seconds. The cyan tick on the whisker is the median.
Spread against the cohort
p10 76.0 s / median 150.6 s / p90 295.5 s
Scale is shared with every product in the cohort, so the bars are directly comparable between pages. Ratio of p90 to median is 1.96x.
Outcome split
30 pass / 12 partial / 3 fail
SEO and GEO measurements
Taken from the public marketing page this product generated in prompt 8, fetched with JavaScript disabled.
HTML bytes, no JS
6.0 KB
6,144 bytes of real text content with scripts blocked
LCP
2,840 ms
Throttled connection, cold cache
CLS
0.110
Cumulative layout shift on load
SEO and GEO subscore
48.0
0 to 100, worth 12 index points
Raw run data: 45 executions
One row per prompt execution. Nothing is aggregated away, nothing is hidden, including the failures.
| Run | Prompt | Prompt target | Started | Duration (s) | Outcome | Note |
|---|---|---|---|---|---|---|
| R01 | P01 | Tenant and user model | 11 Aug 2026 08:00 UTC | 85.5 | pass | |
| R01 | P02 | Authentication and session | 11 Aug 2026 08:03 UTC | 70.9 | pass | |
| R01 | P03 | Projects CRUD with tenancy | 11 Aug 2026 08:06 UTC | 120.6 | pass | |
| R01 | P04 | Public REST API with keys | 11 Aug 2026 08:10 UTC | 407.6 | partial | Pagination off by one at the final page boundary. |
| R01 | P05 | Outbound webhooks with signatures | 11 Aug 2026 08:19 UTC | 186.3 | partial | Outbound calls are constrained by the runtime. Delivery succeeded through a gateway, signing was added by hand. |
| R01 | P06 | Background jobs | 11 Aug 2026 08:24 UTC | 207.2 | pass | |
| R01 | P07 | Admin panel | 11 Aug 2026 08:29 UTC | 114.7 | pass | |
| R01 | P08 | Public marketing surface | 11 Aug 2026 08:33 UTC | 91.1 | pass | |
| R01 | P09 | Hardening pass | 11 Aug 2026 08:37 UTC | 197.2 | pass | |
| R02 | P01 | Tenant and user model | 12 Aug 2026 08:00 UTC | 53.3 | pass | |
| R02 | P02 | Authentication and session | 12 Aug 2026 08:02 UTC | 147.1 | partial | Error envelope inconsistent between two of the endpoints. |
| R02 | P03 | Projects CRUD with tenancy | 12 Aug 2026 08:07 UTC | 149.7 | pass | |
| R02 | P04 | Public REST API with keys | 12 Aug 2026 08:11 UTC | 265.9 | pass | |
| R02 | P05 | Outbound webhooks with signatures | 12 Aug 2026 08:18 UTC | 228.2 | partial | Outbound calls are constrained by the runtime. Delivery succeeded through a gateway, signing was added by hand. |
| R02 | P06 | Background jobs | 12 Aug 2026 08:24 UTC | 371.2 | pass | |
| R02 | P07 | Admin panel | 12 Aug 2026 08:32 UTC | 162.5 | pass | |
| R02 | P08 | Public marketing surface | 12 Aug 2026 08:36 UTC | 103.5 | partial | Validation added on the write path only. Read path left unguarded. |
| R02 | P09 | Hardening pass | 12 Aug 2026 08:40 UTC | 129.1 | partial | Crawlable page test scored low: the deployed front end renders entirely on the client. |
| R03 | P01 | Tenant and user model | 13 Aug 2026 08:00 UTC | 93.7 | pass | |
| R03 | P02 | Authentication and session | 13 Aug 2026 08:03 UTC | 131.6 | pass | |
| R03 | P03 | Projects CRUD with tenancy | 13 Aug 2026 08:07 UTC | 185.2 | pass | |
| R03 | P04 | Public REST API with keys | 13 Aug 2026 08:12 UTC | 442.1 | pass | |
| R03 | P05 | Outbound webhooks with signatures | 13 Aug 2026 08:22 UTC | 74.9 | partial | Outbound calls are constrained by the runtime. Delivery succeeded through a gateway, signing was added by hand. |
| R03 | P06 | Background jobs | 13 Aug 2026 08:25 UTC | 163.5 | partial | Delivered the main path, missed the boundary case named in the pass criteria. |
| R03 | P07 | Admin panel | 13 Aug 2026 08:30 UTC | 138.1 | pass | |
| R03 | P08 | Public marketing surface | 13 Aug 2026 08:34 UTC | 77.6 | partial | Pagination off by one at the final page boundary. |
| R03 | P09 | Hardening pass | 13 Aug 2026 08:37 UTC | 55.8 | pass | |
| R04 | P01 | Tenant and user model | 14 Aug 2026 08:00 UTC | 64.6 | pass | |
| R04 | P02 | Authentication and session | 14 Aug 2026 08:03 UTC | 153.5 | fail | Agent looped on the same edit for the full timeout window of 20 minutes. |
| R04 | P03 | Projects CRUD with tenancy | 14 Aug 2026 08:07 UTC | 135.8 | pass | |
| R04 | P04 | Public REST API with keys | 14 Aug 2026 08:11 UTC | 150.6 | pass | |
| R04 | P05 | Outbound webhooks with signatures | 14 Aug 2026 08:16 UTC | 154.8 | partial | Outbound calls are constrained by the runtime. Delivery succeeded through a gateway, signing was added by hand. |
| R04 | P06 | Background jobs | 14 Aug 2026 08:20 UTC | 170.9 | fail | Task abandoned by the agent with no artifact produced. |
| R04 | P07 | Admin panel | 14 Aug 2026 08:25 UTC | 224.4 | pass | |
| R04 | P08 | Public marketing surface | 14 Aug 2026 08:31 UTC | 204.8 | pass | |
| R04 | P09 | Hardening pass | 14 Aug 2026 08:36 UTC | 78.7 | partial | Crawlable page test scored low: the deployed front end renders entirely on the client. |
| R05 | P01 | Tenant and user model | 15 Aug 2026 08:00 UTC | 106.0 | pass | |
| R05 | P02 | Authentication and session | 15 Aug 2026 08:03 UTC | 161.4 | pass | |
| R05 | P03 | Projects CRUD with tenancy | 15 Aug 2026 08:08 UTC | 184.9 | pass | |
| R05 | P04 | Public REST API with keys | 15 Aug 2026 08:13 UTC | 193.6 | pass | |
| R05 | P05 | Outbound webhooks with signatures | 15 Aug 2026 08:18 UTC | 124.0 | partial | Outbound calls are constrained by the runtime. Delivery succeeded through a gateway, signing was added by hand. |
| R05 | P06 | Background jobs | 15 Aug 2026 08:22 UTC | 378.7 | pass | |
| R05 | P07 | Admin panel | 15 Aug 2026 08:31 UTC | 315.2 | pass | |
| R05 | P08 | Public marketing surface | 15 Aug 2026 08:38 UTC | 191.8 | fail | Generated migration conflicted with the existing schema and was not reconciled. |
| R05 | P09 | Hardening pass | 15 Aug 2026 08:43 UTC | 100.8 | pass |
Pricing
No subscription. The deployed application pays for its own compute and storage from a prepaid balance.
- Entry price
- EUR 0.00
- Value subscore
- 76.0
- Index points per euro
- 60.10
Capabilities as measured
- stack
- Canister backend with a generated web front end
- deploy
- Deployed to a decentralised runtime
- database
- On chain persistent storage
- auth
- Cryptographic identity
- background jobs
- Timers, no external queue
- webhooks
- Outbound is constrained by the runtime
- rest api
- Gateway exposed
- mcp server
- No
- export code
- Yes, source is portable
- custom domain
- Yes
- ssr
- No
Screenshot
Caffeine landing page, captured by the lab on 14 Aug 2026 at 1440 by 900. Stored locally, never hotlinked.
Caffeine, captured 14 Aug 2026.
Sources and fetch dates
Every claim on this page traces to one of these.
| Source | URL | Fetched |
|---|---|---|
| Caffeine pricing page | https://caffeine.ai/pricing | 14 Aug 2026 |
| Caffeine product site | https://caffeine.ai | 14 Aug 2026 |
| LLM Tier run harness output | https://www.llm-tier.com/api/runs.json | 14 Aug 2026 |
Rich results carry this composite as 3.0 on a 0 to 5 scale because that is the only scale the format accepts. The number on this page, 0 to 100, is the real one.
Reader index
Voted by readers with an account, one score per person per axis. Published beside the lab index, never inside it.
Reader index
not rated yet
Lab index
60.10
No reader has scored Caffeine yet. The first 3 voters put the column on the board.
The reader index is a weighted mean of reader submitted axis scores using the published index weights, renormalised over the axes readers have scored. It is never blended into the lab index. How this works.
Open rating mean
n/a
0 published open ratings, submitted without an account
Lab index
60.10
Measured, not voted
Gap to the measurement
n/a
Positive means open raters put it above the measurement
What agent performance looked like in practice
The only product in the index that deploys to a decentralised runtime, which makes it the hardest to compare fairly. Data durability and tenancy were strong, and the application kept running with no server to maintain. Everything that assumes a conventional web stack, from the crawlable page test to the outbound webhook prompt, either scored low or did not apply, and the first paint numbers are the worst in the cohort.
Across 45 recorded executions this product met the published pass criterion cleanly 30 times, partially 12 times and failed 3 times. The rows above carry the reviewer note for every non clean outcome, which is the only way to tell a product that fails loudly from one that fails quietly.