Augment Code logo

coding-agent / rank 7 of 41 / cycle 04

Augment Code benchmark

Indexes a whole repository before it acts, and on a large codebase that is worth the wait: it found the existing tenancy helper in prompt 3 and reused it rather than writing a second one, which no other agent did in every run. The price is the highest in the agent group and the value subscore reflects it. On a small project the advantage largely disappears.

Last run: 14 Aug 2026 17:40 UTCNext scheduled run: 21 Aug 2026 09:00 UTCOfficial siteAugment Code reviews and ratings

Composite index

75.92

of 100 | 0 versus cycle 03

At a glance

Composite index

75.92

rank 7 of 41

Rank

#7

of 41 measured builders

Median speed

71.6 s

p10 44.9 s, p90 110.0 s

Reliability

88.9%

strict pass rate

SEO score

58.0

0 to 100, SEO and GEO axis

Three strongest axes

  • Code ownership97.0
  • Reliability88.9
  • Agent performance86.0

Three weakest axes

  • Speed43.9
  • SEO and GEO58.0
  • Design output62.0

Median per prompt

71.6 s

p10 44.9 s, p90 110.0 s

Strict pass rate

88.9%

40 pass, 3 partial, 2 fail

Executions recorded

45

5 runs of 9 prompts

Entry price

EUR 45.00

Cheapest plan that can finish the specification

Per axis subscores

Composite recomputed from these ten rows: 75.93. That is the arithmetic behind the index, and you can repeat it by hand.

Per axis subscores against the roster median, and weighted contribution
AxisSubscore 0 to 100, tick is the roster medianScoreMedianWeight %Contribution
Agent performance86.075.01815.48
Reliability88.971.11816.00
Scalability84.070.01310.92
SEO and GEO58.058.0126.96
API and MCP74.062.096.66
Integrations76.066.086.08
Design output62.071.074.34
Speed43.933.073.07
Value70.076.053.50
Code ownership97.090.032.91

Speed distribution

Every recorded prompt duration for Augment Code, in seconds. The cyan tick on the whisker is the median.

min 33.8 smax 157.4 s

Spread against the cohort

p10 44.9 s / median 71.6 s / p90 110.0 s

Scale is shared with every product in the cohort, so the bars are directly comparable between pages. Ratio of p90 to median is 1.54x.

Outcome split

40 pass / 3 partial / 2 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

22.0 KB

22,528 bytes of real text content with scripts blocked

LCP

1,800 ms

Throttled connection, cold cache

CLS

0.040

Cumulative layout shift on load

SEO and GEO subscore

58.0

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.

Raw per run measurements
RunPromptPrompt targetStartedDuration (s)OutcomeNote
R01P01Tenant and user model11 Aug 2026 08:00 UTC97.0pass
R01P02Authentication and session11 Aug 2026 08:03 UTC99.2pass
R01P03Projects CRUD with tenancy11 Aug 2026 08:07 UTC134.7passReused the existing tenancy helper instead of writing a second one. No duplicate policy code in the diff.
R01P04Public REST API with keys11 Aug 2026 08:11 UTC83.8pass
R01P05Outbound webhooks with signatures11 Aug 2026 08:14 UTC65.1pass
R01P06Background jobs11 Aug 2026 08:17 UTC70.8failDeployed URL returned 500 on every request.
R01P07Admin panel11 Aug 2026 08:21 UTC89.5pass
R01P08Public marketing surface11 Aug 2026 08:24 UTC82.4partialError envelope inconsistent between two of the endpoints.
R01P09Hardening pass11 Aug 2026 08:28 UTC46.0pass
R02P01Tenant and user model12 Aug 2026 08:00 UTC34.3pass
R02P02Authentication and session12 Aug 2026 08:02 UTC71.6pass
R02P03Projects CRUD with tenancy12 Aug 2026 08:05 UTC87.9passReused the existing tenancy helper instead of writing a second one. No duplicate policy code in the diff.
R02P04Public REST API with keys12 Aug 2026 08:09 UTC58.6pass
R02P05Outbound webhooks with signatures12 Aug 2026 08:12 UTC48.8pass
R02P06Background jobs12 Aug 2026 08:15 UTC51.1pass
R02P07Admin panel12 Aug 2026 08:17 UTC135.1pass
R02P08Public marketing surface12 Aug 2026 08:22 UTC51.2pass
R02P09Hardening pass12 Aug 2026 08:24 UTC59.2pass
R03P01Tenant and user model13 Aug 2026 08:00 UTC46.7pass
R03P02Authentication and session13 Aug 2026 08:02 UTC103.6pass
R03P03Projects CRUD with tenancy13 Aug 2026 08:06 UTC112.8passReused the existing tenancy helper instead of writing a second one. No duplicate policy code in the diff.
R03P04Public REST API with keys13 Aug 2026 08:10 UTC49.9pass
R03P05Outbound webhooks with signatures13 Aug 2026 08:13 UTC44.3partialPagination off by one at the final page boundary.
R03P06Background jobs13 Aug 2026 08:15 UTC82.9partialDelivered the main path, missed the boundary case named in the pass criteria.
R03P07Admin panel13 Aug 2026 08:19 UTC48.8pass
R03P08Public marketing surface13 Aug 2026 08:22 UTC72.3pass
R03P09Hardening pass13 Aug 2026 08:25 UTC84.1pass
R04P01Tenant and user model14 Aug 2026 08:00 UTC36.1pass
R04P02Authentication and session14 Aug 2026 08:02 UTC57.6pass
R04P03Projects CRUD with tenancy14 Aug 2026 08:05 UTC51.0passReused the existing tenancy helper instead of writing a second one. No duplicate policy code in the diff.
R04P04Public REST API with keys14 Aug 2026 08:08 UTC86.6pass
R04P05Outbound webhooks with signatures14 Aug 2026 08:11 UTC67.1failTask abandoned by the agent with no artifact produced.
R04P06Background jobs14 Aug 2026 08:14 UTC84.5pass
R04P07Admin panel14 Aug 2026 08:18 UTC157.4pass
R04P08Public marketing surface14 Aug 2026 08:23 UTC45.8pass
R04P09Hardening pass14 Aug 2026 08:25 UTC112.0pass
R05P01Tenant and user model15 Aug 2026 08:00 UTC106.9pass
R05P02Authentication and session15 Aug 2026 08:03 UTC68.4pass
R05P03Projects CRUD with tenancy15 Aug 2026 08:06 UTC74.0passReused the existing tenancy helper instead of writing a second one. No duplicate policy code in the diff.
R05P04Public REST API with keys15 Aug 2026 08:10 UTC80.6pass
R05P05Outbound webhooks with signatures15 Aug 2026 08:13 UTC33.8pass
R05P06Background jobs15 Aug 2026 08:16 UTC83.4pass
R05P07Admin panel15 Aug 2026 08:19 UTC103.0pass
R05P08Public marketing surface15 Aug 2026 08:23 UTC34.8pass
R05P09Hardening pass15 Aug 2026 08:25 UTC49.3pass

Pricing

From EUR 45 per seat per month. There is a trial but no free tier of any size.

Entry price
EUR 45.00
Value subscore
70.0
Index points per euro
1.69

Capabilities as measured

stack
Your repository, any framework
deploy
Not included
database
Whatever you already run
auth
It reuses what your repository already has
background jobs
Implemented in your stack
webhooks
Signed with a delivery log
rest api
Written to your existing conventions
mcp server
Yes
export code
The code was never anywhere else
custom domain
Your hosting decides
ssr
Your framework decides

Screenshot

Augment Code landing page, captured by the lab on 14 Aug 2026 at 1440 by 900. Stored locally, never hotlinked.

Augment Code landing page, captured by the lab at 1440 by 900

Augment Code, captured 14 Aug 2026.

Sources and fetch dates

Every claim on this page traces to one of these.

Sources with fetch dates
SourceURLFetched
Augment Code pricing pagehttps://augmentcode.com/pricing14 Aug 2026
Augment Code product sitehttps://augmentcode.com14 Aug 2026
LLM Tier run harness outputhttps://www.llm-tier.com/api/runs.json14 Aug 2026

Rich results carry this composite as 3.8 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.

Augment Code head to head

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

75.92

2 readers have scored 100% of the index weight so far. A reader index is published at 3 voters and 50% coverage.

Reader mean against the measured lab subscore for each axis
AxisWeightLabReadersVotesReader mean
Agent performance18%86.076.52
Reliability18%88.985.52
Scalability13%84.074.52
SEO and GEO12%58.053.02
API and MCP9%74.069.52
Integrations8%76.071.52
Design output7%62.062.01
Speed7%43.940.02
Value5%70.051.01
Code ownership3%97.093.52

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

75.92

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

Indexes a whole repository before it acts, and on a large codebase that is worth the wait: it found the existing tenancy helper in prompt 3 and reused it rather than writing a second one, which no other agent did in every run. The price is the highest in the agent group and the value subscore reflects it. On a small project the advantage largely disappears.

Across 45 recorded executions this product met the published pass criterion cleanly 40 times, partially 3 times and failed 2 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.