Canonical: https://weaveos.com/research/q3-2026

[Skip to the report](https://weaveos.com/research/q3-2026#research-report-main)

Report

Weave AI Impact Report:\
Q3 2026
=======

Published October 2026 | Reporting period: July to September 2026\
System-measured benchmarks from engineering organizations on the Weave platform. No surveys.

[Read the report →](https://weaveos.com/research/q3-2026#executive-summary)

Contents

1. [Executive summary](https://weaveos.com/research/q3-2026#executive-summary)
2. [Introduction](https://weaveos.com/research/q3-2026#report-gate)
3. [Part 1: Weave AI Impact Index](https://weaveos.com/research/q3-2026#report-gate)
4. [Speed: AI output share · Skills](https://weaveos.com/research/q3-2026#report-gate)
5. [Quality: Revert rate · The review economy](https://weaveos.com/research/q3-2026#report-gate)
6. [Cost: Spend vs output · Token inflation · Cohort economics](https://weaveos.com/research/q3-2026#report-gate)
7. [Agents: Where $100 goes · Spirals](https://weaveos.com/research/q3-2026#report-gate)
8. [Part 2: Weave Engineering Index](https://weaveos.com/research/q3-2026#report-gate)
9. [Output trend · Benchmark grid · Work-type ratios](https://weaveos.com/research/q3-2026#report-gate)
10. [Spotlight: PE-backed companies](https://weaveos.com/research/q3-2026#report-gate)
11. [Part 3: Legacy Metric Index](https://weaveos.com/research/q3-2026#report-gate)
12. [PR counts and lines of code · Time to 10th PR](https://weaveos.com/research/q3-2026#report-gate)
13. [Part 4: Prompt Routing Index](https://weaveos.com/research/q3-2026#report-gate)
14. [Router savings · Router 2.0 benchmarks](https://weaveos.com/research/q3-2026#report-gate)
15. [Takeaways](https://weaveos.com/research/q3-2026#report-gate)
16. [Methodology](https://weaveos.com/research/q3-2026#report-gate)
17. [About Weave](https://weaveos.com/research/q3-2026#report-gate)

## Executive summary

21 months into the agentic era, Q2 left one question open: was the gain a one-time step or a slope? Q3 answers it. It is a slope, and it got steeper. Average output per employee rose 87% between June and September, the steepest three-month rise in 21 months of Weave Index data. AI spend rose faster still, and routing is becoming a normal part of the stack.

Scroll sideways or zoom to read the full chart.

Three themes define the quarter:

### 1. The output curve steepened

Platform-wide, average monthly output per employee went from 70 units in June to 89 in July, 104 in August, and 131 in September. July's 28% rise is the largest single-month gain in the series; August added 17% and September 26%. Averaged over the quarter, output per employee ran 72% above the Q2 average. This is not just a change in who is on the platform. The constant-cohort median, which holds the set of organizations fixed, rose 3.1x over four quarters, faster than the blended average's 2.6x over the same period: teams themselves are accelerating. Part 2 has the detail.

### 2. Spend outran output, but cost growth is slowing

Average AI spend per employee rose from $231 a month in June to $575 in September, 2.5x in three months, while output rose 1.9x. The rate of increase is what changed. Spend per employee rose 69% in July, 34% in August, and 11% in September; output per employee rose 28%, 17%, and 26%. In September, output grew faster than spend for the first time this quarter. Of each $100 of agent session spend, $23 ships cleanly, against $26 in Q2 (Q3 excludes one outlier organization).

1.87x

average output per employee in September against June, the steepest three-month rise in the series

2.5x

average AI spend per employee over the same three months, from $231 to $575 a month

+11%

rise in average AI spend per employee in September, against +26% for output

### 3. Routing is becoming normalized

In September, the Weave Router handled about 1.62 million requests across 52 installations and 234,000 agent sessions. It served 82% of them with a different model than the one requested, and routed traffic billed about 44% less than the requested models would have. 53% of input tokens were prompt-cache reads, which is why Router 2.0, released in September, pins sessions to protect warm caches. On Terminal-Bench 4.0 it solved 62.1% of tasks against 60.6% for GPT-6 Astra, at 52% of the cost and 2.2x the speed.

### The bottom line for leaders

Q2's conclusion holds: the gain is real, it is table stakes, and what separates organizations is conversion. Q3 adds urgency. The curve is steepening, so the distance between organizations that measure value and those that measure volume is widening each month. The bill is growing faster than the output, so the case for governing spend is now a budget question rather than a principle.

1.62M

requests handled by the Weave Router in September, across 234,000 agent sessions

82%

of routed requests were served by a different model than the one requested

52%

of GPT-6 Astra's cost per attempt for Router 2.0 on Terminal-Bench 4.0, at a matching solve rate

## Read the full report

Enter your work email to unlock every chart, benchmark and methodology note in the report.

Work email

Unlock the report →

Website

Free with a work email. We'll send you new Weave research when it's published; unsubscribe anytime. [Privacy policy](https://weaveos.com/privacy)
