feat: add quantization to Runware mappings

[smakosh]

Problem

The six Runware provider mappings in the model catalogue had no quantization field, so the serving precision Runware documents for these deployments wasn't surfaced.

Approach

Set quantization on each Runware mapping in packages/models:

Model	Quantization
GLM-5.2 (zai-glm-5-2)	fp4 (NVFP4)
Kimi K2.6 (moonshotai-kimi-k2-6)	fp4 (NVFP4)
DeepSeek V4 Flash (deepseek-v4-flash)	fp8
DeepSeek V4 Pro (deepseek-v4-pro)	fp8
GPT OSS 120B (openai-gpt-oss-120b)	bf16
Gemma 4 31B IT (google-gemma-4-31b)	bf16

NVFP4 is represented as fp4, matching the existing Quantization union and how other NVFP4-served mappings (e.g. Nebius GLM-5.2) are already recorded in the catalogue.

Verification

• pnpm format
• pnpm build — all 17 tasks pass
• Models package specs pass (125 tests)

🤖 Generated with Claude Code

https://claude.ai/code/session_01BQgw9ZZGqCQo1veSusnqH9

---

Generated by Claude Code

Summary by CodeRabbit

• Chores
  • Updated quantization configurations for DeepSeek V4 Pro/Flash, Google Gemma 4, Moonshot Kimi K2.6, OpenAI GPT-OSS-120B, and ZAI GLM-5.2 models to optimize performance.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 00316dd0-d04d-4a84-b2b7-94220ad0e9a6

📥 Commits

Reviewing files that changed from the base of the PR and between 7879237 and 49a5e5a.

📒 Files selected for processing (5)

• packages/models/src/models/deepseek.ts
• packages/models/src/models/google.ts
• packages/models/src/models/moonshot.ts
• packages/models/src/models/openai.ts
• packages/models/src/models/zai.ts

---

Walkthrough

Runware provider configurations now include quantization metadata for DeepSeek V4 Pro, DeepSeek V4 Flash, Gemma 4 31B IT, Kimi K2.6, GPT OSS 120B, and GLM-5.2.

Changes

Runware quantization metadata

Layer / File(s)	Summary
Add model quantization declarations packages/models/src/models/deepseek.ts, packages/models/src/models/google.ts, packages/models/src/models/moonshot.ts, packages/models/src/models/openai.ts, packages/models/src/models/zai.ts	Runware provider entries now declare fp8 for DeepSeek V4 models, bf16 for Gemma 4 31B IT and GPT OSS 120B, and fp4 for Kimi K2.6 and GLM-5.2.

Estimated code review effort: 1 (Trivial) | ~5 minutes

Possibly related PRs

• theopenco/llmgateway#2875: Updates quantization metadata for Runware model entries.
• theopenco/llmgateway#3072: Modifies the DeepSeek V4 Pro and V4 Flash entries.
• theopenco/llmgateway#3131: Modifies the DeepSeek V4 Pro provider configuration.

Suggested reviewers: steebchen

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly summarizes the addition of quantization fields to Runware model mappings.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch claude/runware-models-quant-pde1on

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}