consens.io
Product Watches Topics Questions Benchmark Ask your own question

Tracked question

What models does Amp’s new ChatGPT-linked subscription actually include?

Historical consensus 2026-08-18 Active
Runs Weekly on Tuesday at 09:00 (Europe/Berlin) Last 2026-09-01 09:13 Europe/Berlin Next 2026-09-08 09:00 Europe/Berlin

Movement at this check

Changed since last check

The breakdown of model assignments per routing mode changed significantly. Low mode changed from GPT-5.6 Terra to GLM-5.2 + GPT-5.6 Sol; High mode oracle changed from GPT-5.6 Sol to Claude Fable 5; Review changed from GPT-5.6 Terra to GPT-5.5; and specific routes for Puck/Search and Librarian were added while general routing details were updated.

Direction shift
—
Agreement
+3 pts vs previous check

Agreement over time

90/100
2026-07-28: 90/100 · No material movement 2026-08-04: 90/100 · No material movement 2026-08-11: 64/100 · The new answer provides more specific details about how the GPT-5.6 variants (Terra and Sol) map to different Amp usage modes (Low, Medium, High) and clarifies that non-OpenAI routes like Claude Fable 5 in Ultra mode are excluded. 2026-08-18: 67/100 · The breakdown of model assignments per routing mode changed significantly. Low mode changed from GPT-5.6 Terra to GLM-5.2 + GPT-5.6 Sol; High mode oracle changed from GPT-5.6 Sol to Claude Fable 5; Review changed from GPT-5.6 Terra to GPT-5.5; and specific routes for Puck/Search and Librarian were added while general routing details were updated. 2026-08-25: 75/100 · The exact model routing breakdown substantially changed: Low main agent shifted from GLM-5.2 to GPT-5.6 Terra, High oracle shifted from Claude Fable 5 to GPT-5.6 Sol, and Code review shifted from GPT-5.5 to GPT-5.6 Terra. Additionally, the new answer specifies that Low/Med/High workflows are covered, while Ultra mode uses Claude Fable and Amp credits. 2026-09-01: 90/100 · The response was reformatted from a table to bullet points with slightly streamlined wording, but all core facts, covered models, modes, and exclusions remain identical. View full chart
You are viewing a historical version. Return to current consensus
67 /100
agreement
  • 3 AI models
  • 2 contradictions
  • 9 sources
Consensus OpenAI GPT-5.4 mini
Models consulted
  • OpenAI GPT-5.6 Luna
  • Google Gemini Gemini 3.7 Flash
  • Grok Grok 4.3 · No reasoning

Consensus at this check

Amp’s ChatGPT-linked subscription does not mean you get the full set of ChatGPT/OpenAI models inside Amp. It mainly lets Amp route certain modes to the GPT-5.6 family using your linked ChatGPT entitlement, while Amp’s own plan still handles the app features and usage limits.ampcodeampcode

In practical terms, the linked subscription is used for these Amp routes:

  • Low mode: GLM-5.2 for the main agent, with GPT-5.6 Sol for the oracle.ampcode
  • Medium mode: GPT-5.6 Sol for both main agent and oracle.ampcode
  • High mode: GPT-5.6 Sol for the main agent, with Claude Fable 5 still appearing for the oracle in current routing.ampcode
  • Puck / Search: GPT-5.6 Terra.ampcode
  • Librarian / Compaction: GPT-5.6 Sol.ampcode
  • Thread reading/summarization: GLM-5.2 in some routes, though this can shift with the linked setup.ampcode
  • Review: GPT-5.5.ampcode

So the short answer is: the ChatGPT-linked subscription mainly covers GPT-5.6 Sol/Terra routes inside Amp, not every model Amp can use. In particular, it does not automatically include models like Claude Fable 5 or other non-OpenAI providers used by some Amp modes.ampcodeampcode

One important caveat: Amp’s routing appears to evolve over time, so exact model assignment can vary by mode and update.

Differences

2 notable differences (2 contradictions) across 3 models.

contradiction · minor detail

Whether Low mode replaces GLM-5.2 with GPT-5.6 Terra when ChatGPT is linked.

Low mode agent switches from GLM-5.2 to GPT-5.6 Terra with a linked ChatGPT sub.

Gemini, Grok

“(serving as the main fast agent, replacing the default GLM-5.2).”

Low mode main agent remains GLM-5.2 with GPT-5.6 Sol as oracle.

OpenAI

How to verify: Check Amp's latest modes documentation regarding model routing for Low mode with a linked ChatGPT account.

contradiction · minor detail

Whether High mode oracle uses GPT-5.6 Sol or Claude Fable 5 with a linked ChatGPT sub.

High mode uses GPT-5.6 Sol for both main agent and oracle when ChatGPT is linked.

Gemini, Grok

“the main agent (with extra-high reasoning) and the auxiliary Oracle”

High mode oracle continues to use Claude Fable 5.

OpenAI

How to verify: Check Amp's routing settings for High mode oracle under a linked ChatGPT plan.

Best answerOpenAI

Follow this question

This page is re-checked weekly. Get one e-mail when the AI consensus shifts — no account needed.

Double opt-in · unsubscribe anytime · your address is only used for these updates.

Sources

  1. 1 Subscriptions, At Last - Amp ampcode.com
  2. 2 Modes & Models - Amp ampcode.com
  3. 3 Pricing — Amp ampcode.com
  4. 4 ampcode.com
  5. 5 openai.com
  6. 6 ampcode.com
  7. 7 ampcode.com
  8. 8 ampcode.com
  9. 9 ampcode.com

Position Map

Where the models stand

Each row is one part of the answer. The cards show the distinct positions; the model chips show who supports each one.

0/100 Direction Shift · Stable
Shared conclusion

Amp’s ChatGPT-linked subscription covers these OpenAI models for the specified modes:

Shared position

Supports this conclusion

  • DeepSeek
  • Gemini
  • OpenAI
Shared conclusion

GPT-5.6 Terra — Low mode, thread reading, and code review.

Shared position

Supports this conclusion

  • DeepSeek
  • Gemini
  • OpenAI
Shared conclusion

GPT-5.6 Sol — Medium mode, plus both the main agent and oracle in High mode.

Shared position

Supports this conclusion

  • DeepSeek
  • Gemini
  • OpenAI
Shared conclusion

This is mode-based routing, not access to every model Amp offers.

Shared position

Supports this conclusion

  • DeepSeek
  • OpenAI
See how each model moved across checks
Model position movement by watch date
ModelJul 28Aug 04Aug 11Aug 18Aug 25Sep 01
OpenAI
Gemini
Grok — — —
DeepSeek — — — — —
Same positionChanged position

Cite this answer

consens.io. (2026-08-18). Consensus answer to "What models does Amp’s new ChatGPT-linked subscription actually include?". Models consulted: OpenAI: gpt-5.6-luna, Google Gemini: gemini-3.7-flash, Grok: grok-4.3-no-reasoning. Consensus model: OpenAI. Sources: https://ampcode.com/news/subscriptions, https://ampcode.com/modes, https://ampcode.com/pricing, https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGd3LJW24ZctcgwQxLfzZeJFeI89kcvJIYBiQ9p2BMMxNSDaHc8tHl1FLoDTIPvqSe_uOSwGZpCZMBznSsl3FL8qg0i9g6Pr6OZQSKCjbYUKo8=, https://openai.com/index/gpt-5-6/, https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHDlQoNCNhcp_Fq8w-YNoU_AGruVsOqSzdQKNJUxlsOE4qrEaV0i8YBhe4-AbGt0Sg6DUlj_ehncG4xpIESXl3pCdnLRfgg-hvaOKH1B_Nty-4XzGM1FjpT8pjsnSY=, https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGvoP7Lfo9vgbkOPRSgEF02ortqmruDhW9-2bWc5dgUa7PA-vFahGrHB0RgIfFXfear6QaPYd3Cd7jnh8jf_Jzjo3lSSITv6TuSt_jRZjvT, https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEPWMpTHZj8JYSUTSoRoJ4lIdmMe4f1GE_VyVsVVrvQxxLNPFq7XMxpHTl4V_3aiHO5uIpofMMoKNnLPwJSY29mYTHfyHkn6klYmbHPtW4T5g==, https://ampcode.com/news/a-dial-for-you Retrieved from https://www.consens.io/s/what-models-does-amps-new-chatgpt-linked-subscription-b1kPgI7ycak4oLjC?version=b28db676e0f1b027bc7689e0

Ask your own question

Consensus Watch

Run history

90/100 latest agreement

Since tracking began: The breakdown of which model powers which specific route/mode (e.g., Low mode oracle vs. main agent, Puck, Compaction, Review) is specified in much finer operational detail in the new consensus, though the core conclusion remains that linked ChatGPT primarily covers GPT-5.6 routes and not non-OpenAI models.

View the full agreement chart

Agreement over time

How strongly the models support the same claims. Every point links to its run below.

100 50 0 2026-07-28: 90/100 · No material movement 2026-08-04: 90/100 · No material movement 2026-08-11: 64/100 · The new answer provides more specific details about how the GPT-5.6 variants (Terra and Sol) map to different Amp usage modes (Low, Medium, High) and clarifies that non-OpenAI routes like Claude Fable 5 in Ultra mode are excluded. 2026-08-18: 67/100 · The breakdown of model assignments per routing mode changed significantly. Low mode changed from GPT-5.6 Terra to GLM-5.2 + GPT-5.6 Sol; High mode oracle changed from GPT-5.6 Sol to Claude Fable 5; Review changed from GPT-5.6 Terra to GPT-5.5; and specific routes for Puck/Search and Librarian were added while general routing details were updated. 2026-08-25: 75/100 · The exact model routing breakdown substantially changed: Low main agent shifted from GLM-5.2 to GPT-5.6 Terra, High oracle shifted from Claude Fable 5 to GPT-5.6 Sol, and Code review shifted from GPT-5.5 to GPT-5.6 Terra. Additionally, the new answer specifies that Low/Med/High workflows are covered, while Ultra mode uses Claude Fable and Amp credits. 2026-09-01: 90/100 · The response was reformatted from a table to bullet points with slightly streamlined wording, but all core facts, covered models, modes, and exclusions remain identical. 2026-07-28 2026-09-01

Checks

Newest first. Open any saved result to read the full consensus from that date.

  1. 2026-09-01 Meaningful change
    90/100 agreement

    The response was reformatted from a table to bullet points with slightly streamlined wording, but all core facts, covered models, modes, and exclusions remain identical.

    Open this consensus
  2. 2026-08-25 Meaningful change
    75/100 agreement

    The exact model routing breakdown substantially changed: Low main agent shifted from GLM-5.2 to GPT-5.6 Terra, High oracle shifted from Claude Fable 5 to GPT-5.6 Sol, and Code review shifted from GPT-5.5 to GPT-5.6 Terra. Additionally, the new answer specifies that Low/Med/High workflows are covered, while Ultra mode uses Claude Fable and Amp credits.

    Open this consensus
  3. 2026-08-18 Meaningful change
    67/100 agreement

    The breakdown of model assignments per routing mode changed significantly. Low mode changed from GPT-5.6 Terra to GLM-5.2 + GPT-5.6 Sol; High mode oracle changed from GPT-5.6 Sol to Claude Fable 5; Review changed from GPT-5.6 Terra to GPT-5.5; and specific routes for Puck/Search and Librarian were added while general routing details were updated.

    Open this consensus
  4. 2026-08-11 Meaningful change
    64/100 agreement

    The new answer provides more specific details about how the GPT-5.6 variants (Terra and Sol) map to different Amp usage modes (Low, Medium, High) and clarifies that non-OpenAI routes like Claude Fable 5 in Ultra mode are excluded.

    Open this consensus
  5. 2026-08-04 Stable
    90/100 agreement

    No meaningful movement detected in this check.

    Open this consensus
  6. 2026-07-28 Stable
    90/100 agreement

    No meaningful movement detected in this check.

    Open this consensus

Related questions

  • Is GLM-5.3-Flash actually better than DeepSeek V4 Flash for coding? 6 models compared
  • GPT-6 Leak: Is OpenAI launching GPT-5.7 or GPT-6 in August 2026 with a 1.5M-token context window, a new 10T-scale pretraining foundation, and major agentic reasoning gains? Separate confirmed facts fr 5 models compared
  • Gemini 3.5 Pro Leak: Is Google’s ‘Cappuccino’ AI model really launching on July 17 with a 2M-token context window, advanced coding, AI agents, and deeper reasoning? Separate confirmed facts from rumor 5 models compared
  • Does the Jevons Paradox mean AI coding tools will actually increase global compute demand? 6 models compared

About this tracked question

3 AI models answered this question independently on 2026-08-18. A judge from a different model family then cross-checked the answers, scored how far they agree and flagged where they differ. The question is re-checked weekly, and every earlier version stays on this page.

AI models can make mistakes – verify important information against the sources above.

How consensus works →

Thanks – this page has been reported for review.

© 2026 consens.io
App Topics Questions Model pulse Benchmark Model guide How consensus works About Terms Privacy Imprint