Add LFM2.5-8B-A1B model page (DOC-63)

[jbuchananr]

Summary

Adds the new LFM2.5-8B-A1B model page and integrates it across the docs site per DOC-63.

New file:

• lfm/models/lfm25-8b-a1b.mdx — Full model card with description, specs (8B params / 1.5B active, 128K context, MoE architecture), use-case cards, and Quick Start tabs (Transformers, llama.cpp, vLLM, SGLang) using LFM2.5 sampling parameters.

Updated files:

• lfm/models/text-models.mdx — Added card in the LFM2.5 section with MoE and 128K badges.
• lfm/models/complete-library.mdx — Added row to the model chart; updated 32K context blurb to note the 128K exception.
• lfm/help/faqs.mdx — Updated context length FAQ (128K for 8B, 32K for rest) and model selection to recommend LFM2.5-8B-A1B for highest performance.
• link-snapshot.yaml — New URL added (via pre-commit hook).
• CLAUDE.md — Added sampling params row.
• MODEL-MATRIX.md — Added model entry and updated snippet mapping.

Badge links (HF, GGUF, MLX, ONNX) use placeholder URLs following the LiquidAI/LFM2.5-8B-A1B-* naming convention — will be updated once the HuggingFace repos are published.

Review & Testing Checklist for Human

• Verify the model description and specs match the intended release messaging
• Confirm badge links will resolve once HuggingFace repos are created (currently placeholders following naming convention)
• Run npx mintlify dev and check the model page renders correctly at /lfm/models/lfm25-8b-a1b
• Verify sampling parameters (temp=0.1, top_k=50, rep_penalty=1.05) match the upstream HF model card when it's published

Notes

• Sampling params follow the LFM2.5 family defaults (temp=0.1, top_k=50, repetition_penalty=1.05) per the request to "reference all existing LFM2.5 versioning". Update if the HF model card specifies different values.
• The FAQ now distinguishes 128K context for the 8B model vs 32K for all others.

Link to Devin session: https://app.devin.ai/sessions/35da4ce02f534793928b1b36fc89f5ce
Requested by: @jbuchananr

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment and CI monitoring

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated (UTC)
liquid-docs	🟢 Ready	View Preview	May 27, 2026, 4:46 PM

💡 Tip: Enable Workflows to automatically generate PRs for you.