---
title: "Together AI four serverless models removed on September 14, 2026: 2026-09-14 | Harbyn"
description: "Together AI four serverless models removed on September 14, 2026, effective Sep 14, 2026; replacement: Qwen/Qwen3.5-9B, zai-org/GLM-5.3-Flash."
canonical: "https://harbyn.com/changes/together/2026-09-01-four-serverless-models-removed-september-14-2026"
last-updated: 2026-09-29
---

# Together AI four serverless models removed on September 14, 2026

Markdown version of <https://harbyn.com/changes/together/2026-09-01-four-serverless-models-removed-september-14-2026>.

Retirement - Together AI. high severity. In effect. History: recorded after the fact.

openai/gpt-oss-20b, google/gemma-4-31B-it, thinkingmachines/Inkling-Small and intfloat/multilingual-e5-large-instruct were deprecated and removed from serverless on September 14, 2026. Replacements: Qwen/Qwen3.5-9B for gpt-oss-20b and zai-org/GLM-5.3-Flash for the Gemma and Inkling models; all but the e5 model remain on dedicated endpoints.

The deadline passed on Sep 14, 2026.

## Facts

| Fact | Value |
| --- | --- |
| Vendor | [Together AI](https://harbyn.com/vendors/together.md) |
| Kind | Retirement |
| Severity | high |
| Status | In effect |
| Announced | Sep 1, 2026 |
| Effective | Sep 14, 2026 |
| Added to Harbyn | Sep 29, 2026 |
| Lead time | Recorded after the fact: added 15 days after the deadline |
| Model IDs | `openai/gpt-oss-20b`, `google/gemma-4-31B-it`, `thinkingmachines/Inkling-Small`, `intfloat/multilingual-e5-large-instruct` |
| Replacement | `Qwen/Qwen3.5-9B`, `zai-org/GLM-5.3-Flash` |
| Vendor page | <https://docs.together.ai/docs/changelog> |
| Review | Reviewed by a person, extracted by a deterministic parser |

## Find it in your code

```sh
npx harbyn scan
```

The scanner reads your repository on your machine and reports every file and line that uses the identifiers above, with the replacement. Nothing leaves your machine.

## Sources

Where this event comes from: the page we read, when we read it, and the hash of that snapshot.

- <https://docs.together.ai/docs/changelog/rss.xml> (changelog, fetched 2026-09-29 02:31 UTC, sha256 `10415e14c75c8a486ad7c21bb4c80df09f1a4b5f3fa031a1c9040f4ea4657ce3`)

## Related changes

### More from Together AI

- [H100 80GB dedicated endpoint price cut to $3.99 per hour](https://harbyn.com/changes/together/2026-09-01-h100-dedicated-endpoint-price-cut.md) (Pricing, announced Sep 1, 2026)
- [Together CLI v2.33.2 improvements](https://harbyn.com/changes/together/2026-09-10-preemptible-compute-for-gpu-clusters-preemptible-compute-is-now-in-public-pr-2.md) (Behavior change, announced Sep 10, 2026)
- [Preemptible GPU compute preview; CLI v2.33.2 removes --scale-to-zero-window](https://harbyn.com/changes/together/2026-09-10-preemptible-gpu-compute-and-cli-v2-33-2.md) (Breaking change, announced Sep 10, 2026)
- [lower serverless pricing for Qwen3.7-Max and Qwen3.8-Flash](https://harbyn.com/changes/together/2026-09-22-qwen-serverless-price-cut.md) (Pricing, effective Sep 22, 2026)

### Changes at other vendors

- [OpenRouter: arcee-ai/trinity-large-thinking expire 2026-10-26](https://harbyn.com/changes/openrouter/2026-10-23-arcee-ai-trinity-large-thinking-expires-2026-10-26.md) (effective Oct 26, 2026)
- [Google Cloud Run and Functions: Node.js 20 runtime decommission](https://harbyn.com/changes/google-cloud/2026-10-08-node-js-20-runtime-decommission.md) (effective Oct 30, 2026)
- [Mistral AI: Z.ai GLM 5.2 (zai-glm-5-2) retires on Mistral on October 31, 2026](https://harbyn.com/changes/mistral/2026-09-29-z-ai-glm-5-2-deprecated-and-scheduled-for-retirement.md) (effective Oct 31, 2026)
- [Fireworks AI: DeepSeek V4 Flash 0731 retires from Serverless Training November 2](https://harbyn.com/changes/fireworks/2026-10-18-upcoming-serverless-training-deprecation-deepseek-v4-flash-0731.md) (effective Nov 2, 2026)

[All upcoming deadlines](https://harbyn.com/deadlines.md)