0.129.0: /compact fails with unknown service_tier parameter

Please use /feedback to upload logs and post the session ID here.

@etraut-openai Thanks for the quick response. I uploaded the logs via /feedback. This is the thread id: 019e0615-6eb1-7d83-828e-12fb9246a749

I use codex to debug the issue. Most likely introduced in this PR #21249

Same issue
26.506.31004 (2604)

Can you please try the CLI v0.130.0 or most recent app version? The issue should be fixed there.

@pakrym-oai I can verify that this is fixed in v0.130.0. I'll close this ticket now. Thank you.

I am still receiving this error using CLI v0.130.0 with Azure hosted GPT-5.5

@cammoore54, please report the issue to the folks at Azure so they're aware of the problem.

@cammoore54, please report the issue to the folks at Azure so they're aware of the problem.

Done

@cammoore54 I'm hosting my gpt-5.5 in Azure and have no problem after v0.130.0.

I am still receiving this error using CLI v0.130.0 with Azure hosted GPT-5.5

@cammoore54 Remove "service_tier = "fast"" from your config.toml. Restart codex and then it should work :)

Fast is not supported by Azure anymore.

@Erivess I still have my service_tier settings in my config.toml
service_tier = "fast"

I do not have service_tier in my config.toml yet still get the error

Error running remote compact task: {
"error": {
"message": "Unknown parameter: 'service_tier'.",
"type": "invalid_request_error",
"param": "service_tier",
"code": "unknown_parameter"
}
}

Same: 2ce8ddeb-02eb-4077-8472-a583be5267d8

For the Azure-hosted case where there is no explicit service_tier = "fast" in config.toml, this workaround fixed it locally for me:

[features]
fast_mode = false

Then restart Codex / start a new session and rerun /compact.

Why this seems related: the active provider/model metadata still exposed a fast-speed path, but reported no supported service tiers for that provider, so remote compact could still end up sending service_tier even though I had not configured it directly.

For those of you still getting this issue even after updating. I found i had a default-service-tier = fast entry in .codex-global-state.json (and its backup file). Removing this entry fixed the problem for me.

1. git clone this repo
2. ask codex to drop the service_tier parameter from remote compaction requests
3. recompile (ideally without LTO or fat-LTO)
4. win (and wait for a real fix from OpenAI or Azure)

---

A naive (it disables service_tier even for non-Azure remote compaction) patch can be seen here:
https://github.com/Elefunc/codex/commit/c0581fc0a4da98bff5c3832ddc91fb24679ac8e8

---

You can create an Azure-only script/alias and use it like this

#!/usr/bin/env bash
set -euo pipefail

embedded_azure_openai_api_key='#'
AZURE_OPENAI_API_KEY=${AZURE_OPENAI_API_KEY:-$embedded_azure_openai_api_key}

if [[ -z "${AZURE_OPENAI_API_KEY:-}" ]]; then
  printf 'ca: AZURE_OPENAI_API_KEY is empty\n' >&2
  exit 1
fi

export AZURE_OPENAI_API_KEY

exec /home/cetin/tmp/codex-rs-target/release/codex \
  --dangerously-bypass-approvals-and-sandbox \
  -c model_reasoning_effort=xhigh \
  -c service_tier=fast \
  --enable multi_agent \
  -m gpt-5-5 \
  -c 'model_catalog_json="/home/#/rt/ca.models.json"' \
  -c 'model_provider="azure"' \
  -c 'model_providers.azure.name="Azure"' \
  -c 'model_providers.azure.base_url="https://#.openai.azure.com/openai/v1"' \
  -c 'model_providers.azure.wire_api="responses"' \
  -c 'model_providers.azure.supports_websockets=false' \
  -c 'model_providers.azure.env_http_headers={ "api-key" = "AZURE_OPENAI_API_KEY" }' \
  "$@"

ask Codex to create ca.model.json for you!

0.132.0 still has the same issue!

My patches against 0.132.0 that special-case Azure connections:
https://github.com/Elefunc/codex/commit/71057e6abccd11186674888f7c8bf2991fe834d5

For those of you still getting this issue even after updating. I found i had a default-service-tier = fast entry in .codex-global-state.json (and its backup file). Removing this entry fixed the problem for me.

This worked for me. Note: first close Codex, then remove this entry in the JSON file, save the JSON file, and re-open Codex

My patches against 0.133.0 are at https://github.com/Elefunc/codex/commit/9c37cdf619831eda4c5a11f1e4de6e0c9b207ee4!

My patches against 0.134.0 are at https://github.com/Elefunc/codex/commit/784669bd7182d4006d57da0cbfe4dc9ebb1bfaf6!

My patches against 0.135.0 are at https://github.com/Elefunc/codex/commit/47182c23e840a2c57205e75a583312472b98677a!

My patches against 0.136.0 are at https://github.com/Elefunc/codex/commit/1c6142202bd2147f3bab2bf8e2d65e68340d9e6f!

My patches against 0.137.0 are at https://github.com/Elefunc/codex/commit/0c0d09d84d4c7a58a6aeb0beb221e52d0a72b335!

0.138.0 has resolved the issue for me!