---
schema: 1
type: video
title: We Tested GPT-5.5 for 3 Weeks. It's a Beast.
description: >-
  OpenAI just dropped GPT-5.5—and after three weeks of hands-on testing at
  Every, the headline is its coding ability. On Every's Senior Engineer
  Benchmark, GPT-5.5 scored 62.5 out of 100. That’s about a 30-point leap over
  Claude Opus 4.7. Human senior engineers score in the 80s and 90s, so it's not
  all the way there yet, but it's the first model that's really closing the gap.
  Our team has been running GPT-5.5 on everything: coding, writing, knowledge
  work, and Open Claw. A few things stood out: 1. GPT-5.5 rewrites code from
  first principles instead of patching existing messes 2. It's an absolute beast
  when paired with a contract-style plan written by Claude Opus 4.7 4. It's a
  surprisingly restrained, effective business writer and voice mimic 5. Opus 4.7
  still gets more trust for sharp insight and evaluation, and still wins on vibe
  coding from underspecified prompts Every is the only subscription you need to
  stay at the edge of AI. Subscribe today: https://every.to/subscribe Read the
  full vibe check: https://every.to/p/gpt-5-5 00:00 It's Model Release Day!
  02:04 Vibe Check Roadmap 02:45 The Senior Engineer Benchmark, Explained 04:04
  Why 5.5 Executes Better 05:48 Opus 4.7 Plan Secret Sauce 07:00 Real World
  Build Tests 07:53 Where 5.5 Falls Short 08:50 Language Strengths and
  Weaknesses 09:08 Writing Quality and Voice 09:50 Knowledge Work and the Codex
  Desktop App 11:00 Daily Driver Verdict
canonical: >-
  https://liquidgold.cc/channel/everyinc/explore/video/we-tested-gpt-5-5-for-3-weeks-it-s-a-beast-91fa9c893742
markdown: >-
  https://liquidgold.cc/channel/everyinc/video/we-tested-gpt-5-5-for-3-weeks-it-s-a-beast-91fa9c893742.md
llms: null
parent: 'https://liquidgold.cc/channel/everyinc.md'
channel: everyinc
updated: '2026-10-24T01:03:57.609Z'
language: null
children: []
videoId: GROt1Nd4asY
durationSeconds: 763
publishedAt: '2026-04-23T18:03:38.000Z'
searchChunks:
  - ordinal: 0
    text: null
    startSeconds: 11
    endSeconds: 126
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=11'
    heldReason: rights
  - ordinal: 1
    text: null
    startSeconds: 170
    endSeconds: 181
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=170'
    heldReason: rights
  - ordinal: 2
    text: null
    startSeconds: 181
    endSeconds: 248
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=181'
    heldReason: rights
  - ordinal: 3
    text: null
    startSeconds: 259
    endSeconds: 348
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=259'
    heldReason: rights
  - ordinal: 4
    text: null
    startSeconds: 362
    endSeconds: 427
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=362'
    heldReason: rights
  - ordinal: 5
    text: null
    startSeconds: 427
    endSeconds: 438
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=427'
    heldReason: rights
  - ordinal: 6
    text: null
    startSeconds: 481
    endSeconds: 491
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=481'
    heldReason: rights
  - ordinal: 7
    text: null
    startSeconds: 491
    endSeconds: 533
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=491'
    heldReason: rights
  - ordinal: 8
    text: null
    startSeconds: 533
    endSeconds: 543
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=533'
    heldReason: rights
  - ordinal: 9
    text: null
    startSeconds: 555
    endSeconds: 568
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=555'
    heldReason: rights
  - ordinal: 10
    text: null
    startSeconds: 598
    endSeconds: 608
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=598'
    heldReason: rights
  - ordinal: 11
    text: null
    startSeconds: 608
    endSeconds: 665
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=608'
    heldReason: rights
  - ordinal: 12
    text: null
    startSeconds: 608
    endSeconds: 665
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=608'
    heldReason: rights
  - ordinal: 13
    text: null
    startSeconds: 676
    endSeconds: 763
    watch: 'https://liquidgold.cc/watch/GROt1Nd4asY?t=676'
    heldReason: rights
---


# We Tested GPT-5.5 for 3 Weeks. It's a Beast.

OpenAI just dropped GPT-5.5—and after three weeks of hands-on testing at Every, the headline is its coding ability. On Every's Senior Engineer Benchmark, GPT-5.5 scored 62.5 out of 100. That’s about a 30-point leap over Claude Opus 4.7. Human senior engineers score in the 80s and 90s, so it's not all the way there yet, but it's the first model that's really closing the gap. Our team has been running GPT-5.5 on everything: coding, writing, knowledge work, and Open Claw. A few things stood out: 1. GPT-5.5 rewrites code from first principles instead of patching existing messes 2. It's an absolute beast when paired with a contract-style plan written by Claude Opus 4.7 4. It's a surprisingly restrained, effective business writer and voice mimic 5. Opus 4.7 still gets more trust for sharp insight and evaluation, and still wins on vibe coding from underspecified prompts Every is the only subscription you need to stay at the edge of AI. Subscribe today: https://every.to/subscribe Read the full vibe check: https://every.to/p/gpt-5-5 00:00 It's Model Release Day! 02:04 Vibe Check Roadmap 02:45 The Senior Engineer Benchmark, Explained 04:04 Why 5.5 Executes Better 05:48 Opus 4.7 Plan Secret Sauce 07:00 Real World Build Tests 07:53 Where 5.5 Falls Short 08:50 Language Strengths and Weaknesses 09:08 Writing Quality and Voice 09:50 Knowledge Work and the Codex Desktop App 11:00 Daily Driver Verdict

Canonical: https://liquidgold.cc/channel/everyinc/explore/video/we-tested-gpt-5-5-for-3-weeks-it-s-a-beast-91fa9c893742 · Channel: [Every](https://liquidgold.cc/channel/everyinc.md)

> Search chunk text is held until public-source provenance is verified.

## Document chunks

> Missing: no stored document chunks are available.