Refresh recommended LLM docs#517
Conversation
Update the LLM recommendations with current OpenHands Index-backed cloud and open-weight model guidance, and feature Qwen3.6-35B-A3B on the local LLM page.\n\nCo-authored-by: openhands <openhands@all-hands.dev>
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Workflows to automatically generate PRs for you. |
all-hands-bot
left a comment
all-hands-bot
left a comment
There was a problem hiding this comment.
🟡 Acceptable - Strong documentation update with current model recommendations. Well-structured tables and helpful links to OpenHands Index results. A few minor improvements suggested.
[RISK ASSESSMENT]
- [Overall PR]
⚠️ Risk Assessment: 🟢 LOW
Documentation-only update with no code, API, or functionality changes. Tables are well-structured, links provide transparency, and recommendations are based on OpenHands Index data. Minor concerns about hardware requirement increases and screenshot references, but no blocking issues.
VERDICT:
✅ Worth merging: Documentation is clear, current, and actionable. Minor suggestions can be addressed in follow-up if needed.
KEY INSIGHT:
Excellent use of structured tables and OpenHands Index links to make model selection transparent and data-driven.
Was this automated review useful? React with 👍 or 👎 to this review to help us measure review quality.
Workflow run: https://github.com/OpenHands/docs/actions/runs/26235356816
| - A Mac with Apple Silicon with at least 32GB of RAM | ||
| Running Qwen3.6-35B-A3B requires: | ||
| - A recent GPU with at least 24GB of VRAM for quantized variants, or multiple GPUs for full precision and larger context windows, or | ||
| - A Mac with Apple Silicon with at least 64GB of unified memory for quantized variants |
There was a problem hiding this comment.
🟡 Suggestion: Hardware requirements increased significantly from 12GB to 24GB VRAM. Consider adding a note in the "News" section or as a callout warning to help users with older hardware understand this change upfront.
Example:
<Warning>
Qwen3.6-35B-A3B requires more VRAM than the previous Qwen3-Coder-30B-A3B-Instruct (24GB vs 12GB for quantized variants). If you have limited hardware, consider using a smaller quantized variant or one of the community-reported models mentioned below.
</Warning>| 3. Search for **"Qwen3-Coder-30B-A3B-Instruct"**, confirm you're downloading from the official Qwen publisher, then proceed to download. | ||
| 3. Search for **"Qwen3.6-35B-A3B"**, confirm you're downloading from the official Qwen publisher, then proceed to download. | ||
|
|
||
|  |
There was a problem hiding this comment.
🟡 Suggestion: The screenshot filename references "devstral" but should show Qwen3.6-35B-A3B. Verify that screenshot files match the current model recommendations, or update filenames/alt text to be model-agnostic (e.g., 02_lm_studio_download_model.png).
| | [Kimi-K2.6](https://github.com/OpenHands/openhands-index-results/tree/main/results/Kimi-K2.6) | `openrouter/moonshotai/kimi-k2.6` | 57.1 | Strong open-weight option, especially for coding and information-gathering tasks. | | ||
| | [DeepSeek-V4-Pro](https://github.com/OpenHands/openhands-index-results/tree/main/results/DeepSeek-V4-Pro) | `openrouter/deepseek/deepseek-v4-pro` | 51.3 | Strong coding and test-generation scores; current Index entry covers three benchmarks. | | ||
| | [MiniMax-M2.7](https://github.com/OpenHands/openhands-index-results/tree/main/results/MiniMax-M2.7) | `openrouter/minimax/minimax-m2.7` | 43.4 | Recommended as a lower-cost open-weight option with strong SWE-bench and SWT-bench scores. Also available from MiniMax-compatible OpenAI endpoints as `openai/MiniMax-M2.7`. | | ||
| | [Qwen3.6-35B-A3B](https://huggingface.co/Qwen/Qwen3.6-35B-A3B) | `openai/Qwen3.6-35B-A3B` for local OpenAI-compatible servers, or `openrouter/qwen/qwen3.6-35b-a3b` through OpenRouter | Not yet listed | Recommended local / self-hosted model for OpenHands. It is open-weight, supports a large context window, and is featured in the [local LLM guide](/openhands/usage/llms/local-llms). | |
There was a problem hiding this comment.
🟡 Suggestion: The entry for Qwen3.6-35B-A3B says "Not yet listed" in the OpenHands Index Average column. Consider either:
- Waiting to merge this PR until the model is evaluated and listed in the Index (for consistency with other recommendations), or
- Adding a note explaining why this model is recommended despite not yet being in the Index (e.g., "Recommended based on community testing and agentic capabilities; Index evaluation pending").
This helps users understand the recommendation basis and maintains trust in the OpenHands Index as the primary evaluation source.
4 participants
Summary
Closes #516
Validation
git diff --checkThis PR was created by an AI agent (OpenHands) on behalf of the user.
@neubig can click here to continue refining the PR