[PR #1393] [MERGED] docs: update Ollama's default context length from 2048 to the VRAM-based defaults #1358

New issue

Closed

opened 2026-10-18 14:19:06 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-18 14:19:06 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/open-webui/docs/pull/1393
Author: @silentoplayz
Created: 9/18/2026
Status: ✅ Merged
Merged: 9/18/2026
Merged by: @silentoplayz

Base: main ← Head: docs/ollama-context-default

📝 Commits (1)

d3737ff docs: update Ollama's default context length from 2048 to the VRAM-based defaults

📊 Changes

5 files changed (+11 additions, -11 deletions)

View changed files

📝 docs/faq.mdx (+1 -1)
📝 docs/features/chat-conversations/rag/index.md (+2 -2)
📝 docs/getting-started/quick-start/connect-a-provider/starting-with-ollama.mdx (+2 -2)
📝 docs/troubleshooting/rag.mdx (+5 -5)
📝 docs/troubleshooting/web-search.mdx (+1 -1)

📄 Description

Summary

Five pages state that Ollama defaults to a 2048-token context length. Ollama v0.15.5 (2026-02-03) replaced that fixed default with one picked from available VRAM. The tiers are 4096 tokens below 24 GiB, 32768 tokens from 24 to 48 GiB, and 262144 tokens at 48 GiB and above, and Ollama's context length page documents the same values. This PR replaces the 2048 claim on each page with the current default, links Ollama's page, and keeps every recommendation to raise the value, which still applies on the 4096-token tier. The 2048 pre-fill of Open WebUI's own num_ctx control is unchanged, because that is Open WebUI's default and still true.

Pages touched: docs/faq.mdx, docs/troubleshooting/rag.mdx, docs/troubleshooting/web-search.mdx, docs/features/chat-conversations/rag/index.md, docs/getting-started/quick-start/connect-a-provider/starting-with-ollama.mdx.

None. I found it while checking the claim against Ollama's release notes.

Checklist

I have reviewed the relevant documentation and matched the existing style.
This PR meets Open WebUI's contribution standards: it is accurate, relevant to users, narrowly scoped, maintainable, and not promotional content, advertising, lead generation, SEO placement, or a request to list a product, service, provider, integration, gateway, tool, or company primarily for visibility.
I understand that PRs that do not meet these standards may be closed without review and will not be merged. Repeated, low-quality, off-topic, promotional, or intentionally misleading submissions may result in the contributor being blocked from future participation in Open WebUI repositories.

Notes for reviewers

Sources: the Ollama v0.15.5 release notes, which list the VRAM tiers verbatim; the v0.17.0 release notes, where the macOS and Windows apps adopted the same rule; and https://docs.ollama.com/context-length. I checked the 2048 pre-fill in AdvancedParams.svelte on the current dev branch of Open WebUI before leaving that sentence alone. Text only, no screenshots.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/open-webui/docs/pull/1393 **Author:** [@silentoplayz](https://github.com/silentoplayz) **Created:** 9/18/2026 **Status:** ✅ Merged **Merged:** 9/18/2026 **Merged by:** [@silentoplayz](https://github.com/silentoplayz) **Base:** `main` ← **Head:** `docs/ollama-context-default` --- ### 📝 Commits (1) - [`d3737ff`](https://github.com/open-webui/docs/commit/d3737ffa7955f5434e5bf6c33a43b4d5875d9dc3) docs: update Ollama's default context length from 2048 to the VRAM-based defaults ### 📊 Changes **5 files changed** (+11 additions, -11 deletions) <details> <summary>View changed files</summary> 📝 `docs/faq.mdx` (+1 -1) 📝 `docs/features/chat-conversations/rag/index.md` (+2 -2) 📝 `docs/getting-started/quick-start/connect-a-provider/starting-with-ollama.mdx` (+2 -2) 📝 `docs/troubleshooting/rag.mdx` (+5 -5) 📝 `docs/troubleshooting/web-search.mdx` (+1 -1) </details> ### 📄 Description ## Summary Five pages state that Ollama defaults to a 2048-token context length. Ollama v0.15.5 (2026-02-03) replaced that fixed default with one picked from available VRAM. The tiers are 4096 tokens below 24 GiB, 32768 tokens from 24 to 48 GiB, and 262144 tokens at 48 GiB and above, and Ollama's context length page documents the same values. This PR replaces the 2048 claim on each page with the current default, links Ollama's page, and keeps every recommendation to raise the value, which still applies on the 4096-token tier. The `2048` pre-fill of Open WebUI's own `num_ctx` control is unchanged, because that is Open WebUI's default and still true. Pages touched: `docs/faq.mdx`, `docs/troubleshooting/rag.mdx`, `docs/troubleshooting/web-search.mdx`, `docs/features/chat-conversations/rag/index.md`, `docs/getting-started/quick-start/connect-a-provider/starting-with-ollama.mdx`. ## Related issue or discussion None. I found it while checking the claim against Ollama's release notes. ## Checklist - [x] I have reviewed the relevant documentation and matched the existing style. - [x] This PR meets Open WebUI's contribution standards: it is accurate, relevant to users, narrowly scoped, maintainable, and not promotional content, advertising, lead generation, SEO placement, or a request to list a product, service, provider, integration, gateway, tool, or company primarily for visibility. - [x] I understand that PRs that do not meet these standards may be closed without review and will not be merged. Repeated, low-quality, off-topic, promotional, or intentionally misleading submissions may result in the contributor being blocked from future participation in Open WebUI repositories. ## Notes for reviewers Sources: the Ollama v0.15.5 release notes, which list the VRAM tiers verbatim; the v0.17.0 release notes, where the macOS and Windows apps adopted the same rule; and https://docs.ollama.com/context-length. I checked the `2048` pre-fill in `AdvancedParams.svelte` on the current dev branch of Open WebUI before leaving that sentence alone. Text only, no screenshots. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi

2026-10-18 14:19:06 +00:00